Datasets:

tomekkorbak

/

pii-pile-chunk3-550000-600000

like 0

[ "DNA adducts from chemotherapeutic agents.", "\nThe guiding principle of early work was the hypothesis that the anti-cancer alkylating drugs acted through their ability to cross-link macromolecules essential for cell division. ", "Not long afterwards, DNA was specified as the essential target, and support for the hypothesis came from evidence that the archetypal agent, mustard gas, could link guanine bases in DNA through their N-7 atoms. ", "Quantitative correlations between alkylation of DNA and its inactivation as a template followed, with bacteriophage as a simple test object, showing that the mean lethal dose was close to a single cross-link in the genome. ", "This conclusion applied to either mustard gas or the more recently introduced platinum drugs. ", "Although both inter- and intra-strand cross-links were effective, it was thought that in cells the inter-strand cross-link would, by preventing the separation of the strands necessary for cell division, and by being more difficult to repair, constitute the more effectively lethal lesion. ", "With repair-deficient bacteria, it also emerged that a single cross-link in the genome was lethal, but proficient bacteria could remove about 20 cross-links through excision repair. ", "Mono-7-alkylguanines were not removed and were evidently inert. ", "Thus, only a few percent of the total alkylation products were the most effective lesions. ", "Parallel studies with cultured mammalian cells gave a rather different picture, in that the mean lethal doses of even hypersensitive cell lines were around 20 or more cross-links per genome, about the same as for resistant strains of bacteria. ", "Most cells could withstand several hundreds of cross-links per genome, and although adducts were removed, there was incomplete removal of cross-links. ", "Some, but not all, sensitive cell lines were deficient in excision repair. ", "Methods were devised for measuring the extents of alkylation of DNA in cells of patients treated with chemotherapeutic drugs; these are mainly immunoassays, and were applied generally to peripheral blood leukocytes, although some tumours were studied. ", "Extents of alkylation of leukocyte DNA were generally of the same order as, or rather less than the mean lethal doses of cultured cells of the 'normal' type, but in some reports for cisplatin-treated patients, very wide variability between individuals was found. ", "A positive correlation between adduct levels, and particularly a very minor adduct recognised specifically by one antibody, and favourable therapeutic outcome was discerned, and suggested to have a pharmacogenetic basis. ", "In several instances, extents of alkylation of tumours were significantly higher than the average for leukocytes; for ovarian and a testicular tumour for cisplatin, and for a plasma cell tumour for melphalan. ", "Nevertheless, these favourable examples would not constitute more than three or four mean lethal doses in the tumour cells, assuming that they had the same sensitivity as 'normal' cell lines: the therapeutic effect would of course be much more favourable if the tumour cells resembled 'sensitive' cell lines. ", "This lack of a favourable difference between extents of alkylation in DNA of patients and the mean lethal dose for normal cells was particularly obvious with the methylating drugs dacarbazine and procarbazine. ", "These considerations stress the need for higher extents of alkylation to be achieved in target tumour DNA for successful chemotherapy. ", "One approach is to give a higher overall dose, and to 'rescue' the bone marrow (known from the earliest report on mustard gas to be the most susceptible tissue) by autologous transplantation. ", "The second, which has yet to reach the clinic, is to convert unreactive prodrugs through enzymic activation into alkylating agents specifically in tumours (see Bagshawe, 1994)." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0.004739336492890996, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "Q:\n\nIt's adequate to clean egg shell to cook with it?", "\n\nIf I want to cook natural farm eggs with its shell at low temperature (65C-150F) for 20-30mins and they have some dirt at shell is better to clean it or not?", "\nIf it's better to clean it what's the better way?", "\n\nA:\n\nBasic food safety dictates that you always start with a clean product, no matter what you are using or how you are cooking it. ", " For your eggs, simply rinse with clean water, or place in a bowl of water. ", " A soft cloth or gentle brush could be used to remove dirt.", "\n\nA:\n\nWhen raising eggs, rules for cleaning eggs for sale varies drastically by jurisdiction. ", " In the US for instance, the USDA issues guides, but actual rules are set by each state and many have rules drastically different than the USDA rules. ", "Some for instance do not allow the eggs to touch water at all, while others require it. ", " Having been licensed in two states, and researched it for other, I will avoid some of the contradictory reasoning used by different jurisdictions, and you are interested in your own usage in any case.", "\nRules for cleaning are largely concerned with storage, not immediate usage. ", " When an egg is laid, it gets a layer on the shell called a bloom. ", " Provided the hen is healthy which is part of the goal of getting farm fresh eggs and using the assumption that the small farmer is taking good care of their birds, the bloom contains natural bacteria fighting effects on the shell and is desirable to leave in place. ", " As soon as the egg is washed, the bloom is removed and this minor protection is lost. ", " Any egg that would be classed as nest clean should generally not be washed for storage to be classed as farm fresh. ", " A nest clean egg should little of no manure, and minimal dirty. ", " Sometimes they do, but the majority of the eggs should not. ", " If dirty, they should be cleaned before sale and used much more promptly. ", "30 days from harvest for an unwashed farm eggs should still be as fresh as any egg from the store, while washed, I would personally use within 2 weeks if possible. ", " If more than a couple eggs in a dozen though are dirty, I would consider if that was the correct source for me, the nesting area may not be kept clean enough at least for my use.", "\nNow, for you personal use, yes clean them before use, but preferable is clean them just before use, or at least not long in advance. ", " When cleaning, you want to water to be at least 95F degree and no more than about 105F (35-40C), about body temp and warmer than the egg. ", " Cool water will cool the egg and create a negative pressure which can draw water, and with it contaminants through the porous shell, while warmer with create a positive pressure and keep the water out. ", " Avoid scrubbing as you may push contaminants in, rather, brushing clean. ", " Most recommendations for raising so never soak, and even for home use this is a good guide in most views as well because again, the shell is porous and soaking will allow water through. ", " In general, again, if the eggs is dirty enough to require soaking or scrubbing, well, you are getting eggs that may not be the quality you are looking for by getting farm fresh instead of factory.", "\nIn general, farm fresh it is hopeful you are reducing your risk of things like salmonella contamination, but if you eggs are not nest clean, that does not really give a lot of confidence you are getting what you are after.", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0, 0, 0, 0, 0.013245033112582781, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "Scans available for all items, just message me what you want. ", "See these and many others on our Facebook page http://www.facebook.com/groups/345011612249645/ or check out the web site www.soulmotion.co.uk Text / Phone 07788 215744 deesu@hotmail.co.uk" ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0.016042780748663103 ]

[ "Endoscopic versus surgical therapy for Barrett's esophagus neoplasia.", "\nEsophagectomy has been the traditional therapy for high-grade dysplasia and intramucosal cancer. ", "Though surgery can completely resect the cancer and the affected lymph nodes, it carries significant morbidity and mortality (often exceeds 2%). ", "New developments in endoscopy have provided less-invasive therapies that can also be used to stage tissue invasion of cancer; they include esophageal mucosal resection (EMR) and endoscopic submucosal dissection. ", "Additional endoscopic therapies include photodynamic therapy, radiofrequency ablation (RFA) and argon plasma coagulation. ", "Combining EMR that targets the cancer and RFA that targets the surrounding Barrett's esophagus offers an alternative to the operative approach when there is no lymph node metastasis. ", "Arguments for surgical esophagectomy include concern for missed lymph node metastasis and incomplete endoscopic resection. ", "Based on EMR's high neoplasia eradication rate and its fewer and more manageable complications, EMR, especially when combined with RFA, appears to be a viable alternative to surgery in early submucosal cancers, that is, sm1." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.014492753623188406, 0, 0, 0.0047169811320754715, 0.00819672131147541, 0.00546448087431694, 0, 0.013392857142857142 ]

[ "\nCalculate the remainder when 3288151 is divided by 15509.", "\n243\nCalculate the remainder when 140691 is divided by 140313.", "\n378\nWhat is the remainder when 13051376 is divided by 9632?", "\n16\nCalculate the remainder when 54076 is divided by 3398.", "\n3106\nWhat is the remainder when 49522864 is divided by 181?", "\n178\nCalculate the remainder when 48503586 is divided by 405.", "\n381\nWhat is the remainder when 77985 is divided by 1833?", "\n999\nWhat is the remainder when 1655789 is divided by 1355?", "\n1334\nWhat is the remainder when 7362968 is divided by 4666?", "\n20\nCalculate the remainder when 1507334 is divided by 215332.", "\n10\nWhat is the remainder when 1790717 is divided by 21318?", "\n5\nCalculate the remainder when 187414491 is divided by 100.", "\n91\nWhat is the remainder when 3951213 is divided by 914?", "\n905\nCalculate the remainder when 98005 is divided by 19510.", "\n455\nCalculate the remainder when 43522 is divided by 27.", "\n25\nCalculate the remainder when 171044 is divided by 2666.", "\n420\nCalculate the remainder when 80643 is divided by 3091.", "\n277\nWhat is the remainder when 63710704 is divided by 8?", "\n0\nWhat is the remainder when 2256132 is divided by 58?", "\n48\nWhat is the remainder when 2399445 is divided by 90?", "\n45\nWhat is the remainder when 192425378 is divided by 3?", "\n2\nCalculate the remainder when 38648218 is divided by 1698.", "\n40\nWhat is the remainder when 61917 is divided by 82?", "\n7\nCalculate the remainder when 280449 is divided by 533.", "\n91\nWhat is the remainder when 5820345 is divided by 33070?", "\n25\nWhat is the remainder when 2742933 is divided by 260?", "\n193\nCalculate the remainder when 103604 is divided by 33843.", "\n2075\nCalculate the remainder when 514675 is divided by 3032.", "\n2267\nCalculate the remainder when 847742 is divided by 2013.", "\n269\nWhat is the remainder when 19533452 is divided by 9766714?", "\n24\nWhat is the remainder when 19347045 is divided by 1938?", "\n1929\nWhat is the remainder when 26647 is divided by 3849?", "\n3553\nCalculate the remainder when 3782819 is divided by 9.", "\n2\nCalculate the remainder when 572052029 is divided by 1412.", "\n1409\nCalculate the remainder when 47427177 is divided by 127.", "\n43\nWhat is the remainder when 267633637 is divided by 637?", "\n635\nCalculate the remainder when 2250568 is divided by 42.", "\n40\nWhat is the remainder when 50751 is divided by 4215?", "\n171\nCalculate the remainder when 45448385 is divided by 1009.", "\n1007\nCalculate the remainder when 1452319 is divided by 484106.", "\n1\nCalculate the remainder when 21342479 is divided by 236.", "\n55\nCalculate the remainder when 415499 is divided by 93.", "\n68\nWhat is the remainder when 1391926 is divided by 1391898?", "\n28\nCalculate the remainder when 414276 is divided by 134248.", "\n11532\nCalculate the remainder when 10096322 is divided by 2728.", "\n2722\nWhat is the remainder when 19806951 is divided by 10491?", "\n10434\nWhat is the remainder when 101701 is divided by 2071?", "\n222\nCalculate the remainder when 4007743 is divided by 8.", "\n7\nCalculate the remainder when 1774219 is divided by 406.", "\n405\nCalculate the remainder when 1622346 is divided by 405547.", "\n158\nWhat is the remainder when 5617758 is divided by 552?", "\n54\nCalculate the remainder when 2086427 is divided by 1927.", "\n1413\nWhat is the remainder when 2232671 is divided by 52?", "\n51\nCalculate the remainder when 2235526 is divided by 30.", "\n16\nWhat is the remainder when 680312 is divided by 78?", "\n74\nWhat is the remainder when 11585275 is divided by 143028?", "\n7\nCalculate the remainder when 2799755 is divided by 699933.", "\n23\nWhat is the remainder when 959661 is divided by 1613?", "\n1539\nWhat is the remainder when 48661031 is divided by 55868?", "\n3\nCalculate the remainder when 12860073 is divided by 86.", "\n63\nCalculate the remainder when 293666718 is divided by 18.", "\n12\nCalculate the remainder when 11303385 is divided by 791.", "\n786\nCalculate the remainder when 64136 is divided by 2007.", "\n1919\nCalculate the remainder when 9362701 is divided by 123.", "\n64\nCalculate the remainder when 641464 is divided by 811.", "\n774\nWhat is the remainder when 52440827 is divided by 65?", "\n62\nCalculate the remainder when 222635 is divided by 111317.", "\n1\nWhat is the remainder when 6024210 is divided by 34?", "\n22\nCalculate the remainder when 11899 is divided by 2675.", "\n1199\nCalculate the remainder when 1151797 is divided by 627.", "\n625\nCalculate the remainder when 1164589 is divided by 85.", "\n4\nCalculate the remainder when 5484972 is divided by 5295.", "\n4647\nWhat is the remainder when 100420 is divided by 5906?", "\n18\nCalculate the remainder when 23074 is divided by 2337.", "\n2041\nWhat is the remainder when 9744546 is divided by 9744504?", "\n42\nWhat is the remainder when 436814 is divided by 5022?", "\n4922\nCalculate the remainder when 36589409 is divided by 2843.", "\n2842\nWhat is the remainder when 1248059 is divided by 1539?", "\n1469\nCalculate the remainder when 10797230 is divided by 2739.", "\n92\nWhat is the remainder when 396025 is divided by 3962?", "\n3787\nWhat is the remainder when 6776 is divided by 2228?", "\n92\nCalculate the remainder when 1688383 is divided by 2580.", "\n1063\nCalculate the remainder when 9663465 is divided by 1932679.", "\n70\nWhat is the remainder when 110597 is divided by 22070?", "\n247\nCalculate the remainder when 9704743 is divided by 45.", "\n43\nWhat is the remainder when 12475570 is divided by 3248?", "\n2\nWhat is the remainder when 4982 is divided by 498?", "\n2\nCalculate the remainder when 76089 is divided by 25233.", "\n390\nWhat is the remainder when 2472527 is divided by 24?", "\n23\nCalculate the remainder when 463619 is divided by 8137.", "\n7947\nWhat is the remainder when 72062 is divided by 71489?", "\n573\nWhat is the remainder when 180110 is divided by 72?", "\n38\nCalculate the remainder when 214271 is divided by 768.", "\n767\nCalculate the remainder when 169229084 is divided by 288.", "\n284\nCalculate the remainder when 11184199 is divided by 907.", "\n889\nWhat is the remainder when 129725943 is divided by 428?", "\n427\nWhat is the remainder when 4620741 is divided by 420067?", "\n4\nCalculate the remainder when 50589 is divided by 12636.", "\n45\nWhat is the remainder when 30686238 is divided by 2789648?", "\n110\nCalculate the remainder when 2717491 is divided by 247.", "\n244\nCalculate the remainder when 298059 is divided by 49676.", "\n3\nCalculate the remainder when 48732639 is divided by 16244138.", "\n225\nCalculate the remainder when 7651 is divided by 2448.", "\n307\nCalculate the remainder when 12905185 is divided by 30.", "\n25\nCalculate the remainder when 525854 is divided by 47788.", "\n186\nCalculate the remainder when 2416896 is divided by 402811.", "\n30\nWhat is the remainder when 522340 is divided by 3300?", "\n940\nCalculate the remainder when 1478234 is divided by 1781.", "\n4\nWhat is the remainder when 906882 is divided by 61?", "\n56\nWhat is the remainder when 1425852 is divided by 215?", "\n187\nWhat is the remainder when 444688 is divided by 86?", "\n68\nWhat is the remainder when 12951448 is divided by 143?", "\n81\nWhat is the remainder when 194222 is divided by 136?", "\n14\nCalculate the remainder when 58380514 is divided by 97.", "\n94\nWhat is the remainder when 154977 is divided by 58?", "\n1\nWhat is the remainder when 210018 is divided by 2795?", "\n393\nCalculate the remainder when 211512 is divided by 158.", "\n108\nWhat is the remainder when 13268 is divided by 1046?", "\n716\nCalculate the remainder when 5171706 is divided by 427.", "\n309\nWhat is the remainder when 6338283 is divided by 119590?", "\n13\nCalculate the remainder when 221878 is divided by 44250.", "\n628\nWhat is the remainder when 19037934 is divided by 710?", "\n704\nCalculate the remainder when 1201644 is divided by 3749.", "\n1964\nWhat is the remainder when 378181 is divided by 3318?", "\n3247\nCalculate the remainder when 8661141 is divided by 433057.", "\n1\nWhat is the remainder when 22863771 is divided by 11431868?", "\n35\nWhat is the remainder when 166850554 is divided by 2139?", "\n2137\nCalculate the remainder when 43766 is divided by 372.", "\n242\nCalculate the remainder when 6251483 is divided by 46.", "\n37\nCalculate the remainder when 815822589 is divided by 984.", "\n981\nCalculate the remainder when 105752539 is divided by 457.", "\n454\nWhat is the remainder when 38551368 is divided by 16842?", "\n30\nCalculate the remainder when 977333 is divided by 51366.", "\n1379\nCalculate the remainder when 473753 is divided by 523.", "\n438\nWhat is the remainder when 29658 is divided by 688?", "\n74\nCalculate the remainder when 1315803 is divided by 407.", "\n379\nCalculate the remainder when 273947 is divided by 7266.", "\n5105\nWhat is the remainder when 6110463 is divided by 1149?", "\n81\nCalculate" ]

{ "pile_set_name": "DM Mathematics" }

[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.016129032258064516, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "<?", "xml version=\"1.0\" encoding=\"utf-8\"?", ">\n<set xmlns:android=\"http://schemas.android.com/apk/res/android\"\n android:shareInterpolator=\"false\"\n android:zAdjustment=\"top\">\n <translate\n android:duration=\"@android:integer/config_mediumAnimTime\"\n android:fromXDelta=\"100%p\"\n android:interpolator=\"@android:anim/accelerate_decelerate_interpolator\"\n android:toXDelta=\"0%p\"></translate>\n <alpha\n android:duration=\"@android:integer/config_mediumAnimTime\"\n android:fromAlpha=\"0\"\n android:toAlpha=\"1\" />\n</set>" ]

{ "pile_set_name": "Github" }

[ 0, 0.02857142857142857, 0.015414258188824663 ]

[ "Cellular signal transduction system (CSTS) plays a fundamental role in maintaining homeostasis of a cell, and perturbations of CSTS lead to diseases such as cancers and diabetes. ", "Most of cellular signaling pathways eventually regulate gene expression, thus the latter can be used as a universal bioassay reflecting the state of CSTS. ", "However, the gene expression profile from a cell reflects a mixture of responses to all active signaling pathways, thus it is a challenge to de-convolute the signals embedded in the gene expression data. ", "With advent of high throughput biotechnology, major public databases host the results of millions of gene expression assays, collected under diverse diseases as well as the experimental conditions designed by experimental biologists to probe almost all aspects of CSTS. ", "The wealth of these big data provides unprecedented opportunities to investigate CSTS under physiological and pathological conditions, but it also poses unprecedented challenges: how to reveal signals from convoluted data and turn big data into knowledge. ", "We hypothesize that given a sufficiently large compendium of gene expression data collected under diverse conditions, in which different parts of CSTS are perturbed (either by designed experiments or by diseases), the cellular signals embedded in the gene expression data can be revealed and their organization can inferred using current state-of-the-art deep learning models. ", "In this project, we will compil a comprehensive compendium of human gene expression data and then employ modern deep-learning algorithms and supercomputers to mine the data. ", "We aim to reveal major cellular signals that regulate gene expression under physiological and pathological conditions and to infer the organization of signals in human CSTS. ", "Combined the identified signals with genomic alteration data and drug response data, we aim to further identify pathways underlying disease such as cancers, to use the genomic data to predict drug sensitivity of cancer cell lines, and to predict patient clinical outcomes, in a pathway-centered manner." ]

{ "pile_set_name": "NIH ExPorter" }

[ 0.0111731843575419, 0.0064516129032258064, 0, 0.003703703703703704, 0.00390625, 0.002652519893899204, 0, 0.005747126436781609, 0 ]

[ "Chittagong Hill Tracts (CHT) is a military controlled area, where all the news are filtered by the military and the Bangladeshi government.", "CHT, where blood has shed for decades and hopes were burnt to ashes by the brutes, constitutes of people who want their voice to be heard. ", "We are here to ensure that the voice of these unheard victims in CHT echo around the world despite the Bangladeshi government trying to suppress them in the biased state run media.", "\nOur email : chtnewsupdate(at)gmail.com\n\nPages\n\nFriday, April 22, 2011\n\nHow I escaped from the clutches of death is indescribable: Polen Chakma\n\nApril 21, 2011\n\nPolen Chakma, an NGO activist, (son of Milon Shanti Chakma, of village Babuchara Noapara, Dighinala) is one of the victims of the 17 April communal attacks on Jumma people in Ramgarh, Manikchari and Guimara. ", "Initially, it was reported that he was disappeared.", "\n\nI work with an NGO named IDF. ", "My place of work is at one of its branch offices at Guimara.", "\n\nOn 17 April, when I went to Jaliapara on business the Bengali settlers caught and beat me up with sticks and then threw me into the river.", "\n\nI ran to the other side of the river, but the settlers caught and tortured me again. ", "Later, they locked me inside an empty house.", "\n\nI had cash Taka 3,000 and a mobile phone set worth Taka 4,000. ", "They took them away.", "\n\nI heard them talking about killing me. ", "I also saw them talking with army and police personnel.", "\n\nAt that time, I also heard some common Bengal people saying: “The army is helping us, but why the police are not doing so?”", "\n\nThen 4 – 5 policemen came there and told those Bengalis: “You were supposed to help us, but instead you have become angry with us.”", "\n\nOverhearing the Bengalis talk like this, it seemed to me that it was the army who had instigated the attack.", "\n\nHow I escaped from the clutches of death is indescribable. ", "It is because of my physical strength, wit and faith in Dharma that I have been able to come back.", "\n\nThe moment the settlers tied me I thought of running away. ", "I noticed that of the two doors of the house in which I was locked in, one was locked with keys while the other was shut with the help of a rope.", "\n\nI decided to escape by opening the second door. ", "I waited till dark, and when it was dark I tore off the rope with my teeth and came out.", "\n\nThe settlers were still near the other door.", "\n\nI crossed a field and came to a stream. ", "It was about 7pm. ", "This is how I escaped.", "\n\nThe settlers hit me on every part of my body. ", "I blocked some blows with my hands, but when I was hit in the head bloods came out and I fell almost unconscious.", "\n\nThen the Bengalis struck me on the hands, on the waist and on all over the body.", "\n\nNow I feel so much pain that it has become difficult for me even to stand up and sit down. ", "And one of my hands is almost crippled.", "\n\nAbout Me\n\nChittagong Hill Tracts (CHT), home of Jumma Indigenous people, is a military controlled area, where all the news are filtered by the military and the Bangladeshi government. ", "No foreigners are allowed to enter this military zone without obtaining permission from the Bangladeshi government.", "\nCHT, where blood has shed for decades and hopes were burnt to ashes by the brutes, constitutes of people who want their voice to be heard. ", "We are here to ensure that the voice of these unheard victims in CHT echo around the world despite the Bangladeshi government trying to suppress them in the biased state run media.", "\nIf you have any query or if your have any organization that publishes articles/reports etc. ", "related to CHT please add our email to your mailing list:\nchtnewsupdate(at)gmail.com\nJoin our facebook page:\nhttp://www.facebook.com/chtnewsupdate\nor follow us on twitter:\nhttp://twitter.com/chtnewsupdate\nhttp://www.youtube.com/chtnewsupdate" ]

{ "pile_set_name": "Pile-CC" }

[ 0.014388489208633094, 0.007194244604316547, 0.005555555555555556, 0.013550135501355014, 0, 0.03125, 0.016666666666666666, 0, 0, 0, 0.03076923076923077, 0, 0, 0, 0, 0, 0, 0, 0.01020408163265306, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.010752688172043012, 0, 0, 0.005555555555555556, 0, 0.016597510373443983 ]

[ "Introduction\n============\n\nPrimary lactose intolerance is a widely diffused gastrointestinal disorder in which the organism cannot completely digest the sugar present in milk and its derivatives, leading to symptoms such as flatulence, meteorism, diarrhea, intestinal pain, and abdominal cramps.[@b1-ceg-11-243]--[@b3-ceg-11-243] In particular, lactose intolerance is the most common form of carbohydrate malabsorption and affects people of all ages.[@b4-ceg-11-243]--[@b6-ceg-11-243] The cause of this malabsorption is due to the deficiency of lactase, an enzyme that is normally present in the small intestine. ", "Under normal conditions, after ingestion, lactose is hydrolyzed by lactase in the intestine (the highest activity of this enzyme is in jejunum) to form glucose and galactose. ", "The 2 monosaccharides are then actively absorbed by the intestinal mucosa. ", "In case of lactase deficiency, lactose reaches the colon where it is fermented by the bacterial flora to produce various gases, including hydrogen (H~2~) and methane (CH~4~). ", "These, in part, are absorbed and subsequently eliminated through the lungs.[@b7-ceg-11-243],[@b8-ceg-11-243] Given that the human organism does not produce either H~2~ or CH~4~, when administering lactose in controlled conditions to lactase-deficient patients, the amount of gases measured in the subject's breath is considered a measure of the unabsorbed sugar. ", "These \"Breath Tests,\" after lactose challenge, are considered a standard diagnostic approach to identify lactase deficiency.[@b7-ceg-11-243]--[@b12-ceg-11-243] Lactose malabsorption is diagnosed when breath test values are above a specific threshold value of either H~2~ or CH~4~, or both.[@b9-ceg-11-243],[@b10-ceg-11-243],[@b13-ceg-11-243] A consensus conference back in 2009 had suggested a standardization of the analytical method to be used for lactose breath tests, with a focus mainly on H~2~ intestinal production and a relatively high threshold value for positives, thus privileging specificity over sensitivity.[@b14-ceg-11-243] During the consensus conference, the issue of intestinal variability of H~2~/CH~4~-producing flora was not taken into account.[@b14-ceg-11-243] The 2017 North American consensus conference did not add much to the specific issue, confirming the guidelines.[@b13-ceg-11-243] Nevertheless, a limited sensitivity can lead to missing or delaying lactate malabsorption diagnosis. ", "Some authors have questioned the Rome guidelines and propose to increase the sensitivity of the assay, promoting analytical approaches that take into consideration equally the production of both gases and the high variability of the subject's intestinal microbiome, with an unpredictable proportion of H~2-~producing microbes, versus CH~4~-producing ones.[@b15-ceg-11-243],[@b16-ceg-11-243] The aim of the present study was to compare different analytical approaches of the same lactose breath test in a retrospective study of patients prescribed the test after medical evaluation of gastrointestinal symptoms for a suspected lactose malabsorption condition.", "\n\nThe genetic variability of intestinal lactase is responsible for lactase persistency in the adult and is extremely variable in the general population, ranging from almost 0 in some Asian Countries to high levels found in Northern Europe.[@b4-ceg-11-243],[@b5-ceg-11-243],[@b9-ceg-11-243],[@b17-ceg-11-243],[@b18-ceg-11-243] The lack of lactase persistency is considered responsible for lactose malabsorption. ", "Recently published studies took showed the incidence of lactose malabsorption among Italian population to be 72%.[@b19-ceg-11-243] Another study recently published by Zadro et al[@b20-ceg-11-243] investigated the frequency of lactase non-persistence genotype among Italian population. ", "The results showed 62.3% of lactase nonpersistence genotype, with no significant differences among 3 macroregions of the country (58.6%, 74.1%, and 67.1% detected in North, Center, and South, respectively).", "\n\nOur study is, to date, the first lactose breath test study conducted in North East Italy on a population of patients suspected of lactose malabsorption.", "\n\nMaterials and methods\n=====================\n\nSubjects and sample collection\n------------------------------\n\nDiagnostic data were retrospectively collected, pooled, and analyzed from lactose breath tests of outpatient subjects referring to a community-based laboratory facility in the Veneto Region in Northeast Italy, Data Medica Padova, from January 2014 to June 2016. ", "The facility was part of a group of private health care providers operating under agreement with the public health care system to deliver healthcare to the public. ", "The center was also ISO9001 certified and accredited for excellence in health care services by Accreditation Canada International. ", "The retrospective observational study was carried out according to Good Practice in Clinical Research, approved by the Data Medica Group Scientific Technical Committee, and, following national Italian legislation, approved by the local reference Ethics Committee for Clinical Trials of the Province of Padua (DM 2016/01). ", "All study participants provided informed written consent prior to study enrollment.", "\n\nThe analysis involved 282 subjects whose breath tests results were obtained from the laboratory database and rendered anonymous before elaboration. ", "For each patient, age, sex, self-declared reason for taking the exam (grouped into \"intestinal pain,\" \"meteorism,\" \"nausea,\" \"diarrhea,\" \"no symptoms specified\"), and breath test diagnostic outcome were collected. ", "H~2~ values at 4 hours and H~2~ and CH~4~ peak values were also considered for further analysis.", "\n\nBreath test\n-----------\n\nBefore the test, patients had to fast and follow a series of dietary, pharmacological, and behavioral prescriptions, as indicated by the manufacturer, and on the day of the assay they were asked to fill a questionnaire with identification data and the reasons for the examination. ", "EXPIROlact® H~2~ Breath Test Kit for the determination of lactose intolerance (Sofar SpA, Milan, Italy) was used for the assay, and a standard procedure furnished by the kit manufacturer was followed. ", "Briefly, following the oral administration of 25 g of lactose suspended in 150 mL of tap water, exhaled samples were harvested every 30 minutes for 4 hours and prepared for H~2~ and CH~4~ analysis.[@b10-ceg-11-243] Basal gas levels were subtracted from H~2~ and CH~4~ ppm, and values at 4 hours and peaks were considered for analysis. ", "No sample had peaks of H~2~ values above 20 ppm within 1.5 hours from lactose ingestion, which would have suggested small intestinal bacterial overgrowth[@b13-ceg-11-243] and would have been eliminated from the study.", "\n\nThe analysis was conducted with a Quintron microlizer BreathTracker Analyzer (QuinTron Instrument Company, Inc., Milwaukee, WI, USA), based on solid-state sensors.", "\n\nStatistical analysis\n--------------------\n\nStatistic analyses were performed using GraphPad Prism (v 5.01; Graphpad Inc., La Jolla, CA, USA). ", "Statistical comparisons for subgroups were computed using 1-way analysis of variance followed by post hoc tests for selected comparisons or 2-tailed Student's *t*-test.", "\n\nResults\n=======\n\nA retrospective analysis was performed on a sample of 282 patients who underwent a lactose breath test between 2014 and 2015 in an outpatient laboratory in Northern Italy. ", "Patients with a diagnosis of major gastrointestinal diseases were excluded from this study. ", "Mean age was 35 years, and females were twice the number of males enrolled in the study. ", "Most patients were sent for the test by a general practitioner or a specialist, with suspected hypolactasia potentially linked to 1 or more gastrointestinal symptoms as declared initially by the patients themselves ([Table 1](#t1-ceg-11-243){ref-type=\"table\"} provides a sample description). ", "None of the samples considered for the present study had unacceptable baseline values (≥20 ppm). ", "Two protocols were followed to determine lactose malabsorption. ", "One is based on the maximum ppm values at 4 hours following lactose challenge of either H~2~ or CH~4~ (Threshold \\>20 ppm), and is the method ordinarily used.[@b2-ceg-11-243],[@b10-ceg-11-243] The second approach took into consideration the stoichiometric sum of the ppm values of both H~2~ and CH~4~ (H~2~+2CH~4~) (Threshold \\>18 ppm) in the same time frame following lactose oral administration.[@b4-ceg-11-243],[@b10-ceg-11-243],[@b15-ceg-11-243] Since in some cases peak values did not correspond to those determined at 4 hours, we further collected those values for comparison.", "\n\nAs shown in [Table 2](#t2-ceg-11-243){ref-type=\"table\"} and [Figure 1](#f1-ceg-11-243){ref-type=\"fig\"}, the percentage of positive samples using the first method was 11.4%, with an average value of H~2~ at 4 hours of 11.34 ppm. ", "When using the stoichiometric sum of H~2~ and CH~4~, the number of positives dramatically increases up to 62.8%. ", "Ppm values obtained using the peak values were clearly higher, although nonsignificantly (*p*-value \\>0.05). ", "Nevertheless in the latter case the number of positive subjects increased, more so when using the second method, a difference leading to a significant diagnostic impact.", "\n\nRelevant differences are thus seen between the different analytical approaches, while the same qualitative differences were found when comparing the positive samples from different subgroups of patients ([Tables 3](#t3-ceg-11-243){ref-type=\"table\"}[](#t4-ceg-11-243){ref-type=\"table\"}--[5](#t5-ceg-11-243){ref-type=\"table\"}). ", "When considering 5 different age ranges, all the analytical approaches show the same trends, with a higher value below 20 years, which lowers and increases again reaching top ppm values when considering the higher age range ([Table 3](#t3-ceg-11-243){ref-type=\"table\"}). ", "No differences were seen between male and females in all methods ([Table 4](#t4-ceg-11-243){ref-type=\"table\"}), and the same distribution was also seen for symptomatology among methods ([Tables 5](#t5-ceg-11-243){ref-type=\"table\"} and [6](#t6-ceg-11-243){ref-type=\"table\"}), with no subgroup significantly higher than the others within each method (*p*-value \\>0.05). ", "Hence, results suggest that the different methods have only a difference in sensitivity, and do not differ in qualitative terms.", "\n\nDiscussion\n==========\n\nThe present retrospective observational study was meant to give a contribution to the discussion on the diagnostic application of lactose breath test. ", "In particular, the aim of the study was to establish the frequency of positive lactose breath tests in a group of patients affected by gastrointestinal symptoms who took the test in a local outpatient ambulatory center in Northeastern Italy. ", "Results unexpectedly indicated a rather low frequency of positive tests when the standard analytical methodology was used.[@b9-ceg-11-243] The same results were thus reevaluated using a different approach used by some other authors,[@b15-ceg-11-243] which was characterized by a higher sensitivity and the same level of precision.[@b15-ceg-11-243] This approach led to a much higher proportion of positive samples, as one would have expected in a population of gastrointestinal patients suspected of lactose intolerance. ", "Lactase deficiency is considered a rather frequent condition, though it is known that it is highly variable in the general population and dependent on race and latitude,[@b4-ceg-11-243],[@b18-ceg-11-243] making it difficult to establish fixed rules for diagnostic appropriateness without a precise figure of what to expect in a given geographical area. ", "Recent studies have been conducted on the Italian local population, showing no significant variations in the frequency of lactase nonpersistence genotype among 3 macro-areas of the country.[@b20-ceg-11-243] The values of the frequency of this genetic condition affecting lactose absorption are expected to fall in the range of 60%--70%, according to the studies of Storhaug et al[@b19-ceg-11-243] and Zadro et al.[@b20-ceg-11-243] When we applied the standard analytical methodology, which takes into consideration H~2~ and CH~4~ produced in the intestinal lumen separately, we identified only 11.7% of the patients as positive for the lactose breath test, a figure rather different from what was expected. ", "When we changed the analytical approach, taking into consideration the sum of the 2 bacterial gases produced from lactose by the intestinal flora, the result changed and approached closer to the expected one (62.8%).", "\n\nWhen lactose reaches the intestinal lumen, the first bacterial population encountered produces H~2~ as an end product. ", "If lactose escapes this first level, it can be metabolized by a different microbial population, producing CH~4~ as metabolic end product.[@b15-ceg-11-243] As a consequence, these 2 molecules together, and not separately, account for the total amount of lactose that is not absorbed. ", "When considering them singularly, 2 problems could arise: first of all, variability in the results linked to variability of microbiome in different patients. ", "A second issue is linked to low sensitivity, potentially leading to the identification of a lower amount of lactose malabsorption cases, with important consequences on patients' health and therapeutic options. ", "Conversely, the sum of the 2 molecules, stoichiometrically corrected (since the hydrogen produced from lactose is twice as much as the amount of CH~4~), is not dependent on the microbiome variability, since it includes both types of microbial populations. ", "The first consensus meeting on the subject in 2009 recommended the use of either H~2~ or CH~4~ to determine positivity to the lactose breath test,[@b14-ceg-11-243] but since then others have challenged the decision, while the 2017 consensus did not touch upon the issue.[@b13-ceg-11-243] In particular, our results indicate that the 2 analytical approaches yield 2 different diagnostic outcomes. ", "While the quantitative difference between the 2 tests is very high, the 2 methods do not seem to carry qualitative differences, at least in the parameters considered in the present retrospective study. ", "Considering different age groups, the frequency trends are similar, and there are no significant differences in the sex stratification of the positive samples. ", "Similarly, we could not find significant differences in the stratification of the groups according to the symptoms declared by the patients.", "\n\nConclusion\n==========\n\nOur results suggest that the analytical methodology used and the threshold for lactose breath tests should be reconsidered, especially in view of the extreme variability known to affect intestinal microbiome, a condition that can greatly affect the diagnostic outcome. ", "Taking into account previous studies, indicating a significant underestimation of lactase genetic nonpersistence when using H~2~/CH~4~ standard breath tests,[@b11-ceg-11-243] our results can contribute to the discussion on the subject that has relevant consequences on lactose malabsorption diagnosis that might be missed.", "\n\nAvailability of data and materials\n==================================\n\nThe data that support the findings of this study are available from Data Medica Group, Padova, but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. ", "Data are however available from the authors upon reasonable request and with permission of Data Medica Group, Padova.", "\n\nWe thank the laboratory staff in Data Medica for helping with data collection, in particular Dr Giovanni Lenzo and Dr Alessia Caltabiano.", "\n\n**Disclosure**\n\nThe authors report no conflicts of interest in this work.", "\n\n{#f1-ceg-11-243}\n\n###### \n\nCharacteristics of the subjects enrolled in the study\n\n Subjects' characteristics \n ------------------------------------------------------- ---------------\n Number of subjects 282\n Males 93\n Females 189\n Age (years)[a](#tfn1-ceg-11-243){ref-type=\"table-fn\"} 35±17\n Lactose intolerance (self-declared) 40/282 (14%)\n Abdominal discomfort 110/282 (39%)\n Diarrhea 71/282 (25%)\n Meteorism 109/282 (39%)\n Nausea 15/282 (5%)\n 2 concomitant symptoms 64/282 (23%)\n 3 or more symptoms 21/282 (7%)\n\n**Note:**\n\nData presented as mean ± SD.", "\n\n###### \n\nBreath test results\n\n Variable Average (ppm) Range (ppm) Number of positive samples (%) Average ppm of positive samples\n ------------------------------------------------------------------ --------------- ------------- -------------------------------- ---------------------------------\n H~2~ (4 hours)[a](#tfn2-ceg-11-243){ref-type=\"table-fn\"} 11.34±10.00 1--65 32 (11.4%) 34.60±8.00\n H~2~ (*C*~max~)[a](#tfn2-ceg-11-243){ref-type=\"table-fn\"} 13.40±9.20 2--65 33 (11.7%) 34.30±8.10\n H~2~+2CH~4~ (4 hours)[b](#tfn3-ceg-11-243){ref-type=\"table-fn\"} 21.40±15.10 3--93 139 (49.3%) 32.80±13.60\n H~2~+2CH~4~ (*C*~max~)[b](#tfn3-ceg-11-243){ref-type=\"table-fn\"} 24.15±14.13 4--93 177 (62.8%) 35.80±14.40\n\n**Notes:**\n\nThreshold \\>20 ppm;\n\nthreshold \\>18 ppm. ", "Data presented as mean ± SD unless otherwise stated.", "\n\n**Abbreviation:** *C*~max~, peak concentration.", "\n\n###### \n\n\\% of positives and mean ppm values among different age groups\n\n Age (years) \\% Positive (average ppm ± SD) \n ------------- -------------------------------- ------------------ ------------------ ------------------\n \\<20 15.6 (33.1±5.0) 16 (32.1±5.6) 51.1 (37.0±14.0) 64.4 (34.5±13.1)\n 20--40 11.3 (35.5±10.6) 11.3 (35.5±10.7) 48.1 (32.4±15.9) 59.3 (31.1±15.0)\n 40--60 11.0 (34.3±4.5) 11.0 (34.3±4.8) 53.7 (31.0±12.2) 68.3 (30.6±11.2)\n \\>60 0 (−) 0 (−) 30.0 (23.3±5.5) 50.0 (24.8±3.9)\n\n**Notes:**\n\nThreshold \\>20 ppm;\n\nthreshold \\>18 ppm.", "\n\n**Abbreviation:** *C*~max~, peak concentration.", "\n\n###### \n\n\\% of positives and mean ppm values between the sexes\n\n Sex \\% Positive (average ppm ± SD) \n --------- -------------------------------- ----------------- ------------------ ------------------\n Males 11.9 (11.2±9.6) 11.9 (13.3±8.9) 51.1 (21.8±14.9) 59.8 (24.5±14.1)\n Females 10.6 (11.4±10.2) 11.2 (13.5±9.4) 50.0 (21.1±15.2) 64.4 (24.0±14.2)\n\n**Notes:**\n\nThreshold \\>20 ppm;\n\nthreshold \\>18 ppm.", "\n\n**Abbreviation:** *C*~max~, peak concentration.", "\n\n###### \n\n\\% of positives and mean ppm values among patients grouped by common symptoms (H~2~ test results)\n\n Symptoms Total H~2~ (4 hours) H~2~ (*C*~max~) \n ----------------- ------- ---------------- ----------------- ----------- ---- ------ -----------\n Intestinal pain 102 15 15.2 36.6±6.4 15 15.2 36.6±6.3\n Meteorism 90 11 12.3 33.2±5.1 11 12.3 33.1±5.0\n Nausea 15 2 7.1 30.7±6.6 2 7.2 31.2±6.7\n Diarrhea 72 10 14.1 38.4±10.2 10 14.2 38.2±11.5\n No symptoms 84 8 10.6 34.3±6.3 9 11.1 32.7±6.7\n Whole sample 32 1.4 33 11.7 \n\n**Note:** Data presented as mean ± SD unless otherwise stated.", "\n\n**Abbreviation:** *C*~max~, peak concentration.", "\n\n###### \n\n\\% of positives and mean ppm values among patients grouped by common symptoms (H~2~+2CH~4~ test results)\n\n Symptoms Total H~2~+2CH~4~ (4 hours) H~2~+2CH~4~ (*C*~max~) \n ----------------- ------- ----------------------- ------------------------ ----------- ------- ------ -----------\n Intestinal pain 102 53 52.1 33.2±15.0 66 65.0 33.7±13.3\n Meteorism 90 41 46.4 33.2±13.7 50 56.0 32.2±12.3\n Nausea 15 8 53.4 32.4±9.5 10 67.4 32.1±8.1\n Diarrhea 72 32 44.2 36.1±19.0 45 63.6 33.3±17.6\n No symptoms 84 48 57.0 30.7±13.2 59 70.1 30.2±12.0\n Whole sample 139 49.3 177.0 62.8 \n\n**Note:** Data presented as mean ± SD unless otherwise stated.", "\n\n**Abbreviation:** *C*~max~, peak concentration.", "\n" ]

{ "pile_set_name": "PubMed Central" }

[ 0.0065252854812398045, 0, 0, 0.005714285714285714, 0.008264462809917356, 0.007897334649555774, 0.00303951367781155, 0.012165450121654502, 0.010526315789473684, 0, 0, 0.002688172043010753, 0, 0.007633587786259542, 0.006211180124223602, 0, 0, 0, 0.020833333333333332, 0, 0, 0.005970149253731343, 0.009216589861751152, 0.01818181818181818, 0.013888888888888888, 0, 0, 0, 0, 0, 0, 0, 0.012027491408934709, 0.004347826086956522, 0.008849557522123894, 0, 0, 0, 0, 0, 0, 0, 0, 0.005758157389635317, 0.0056657223796034, 0.007072135785007072, 0, 0.008264462809917356, 0.0035335689045936395, 0, 0, 0, 0.007575757575757576, 0, 0, 0, 0, 0.006211180124223602, 0.006369426751592357, 0.008547008547008548, 0.02158273381294964, 0, 0, 0.0007564296520423601, 0.0030303030303030303, 0, 0, 0.0027137042062415195, 0, 0.001996007984031936, 0, 0.0032292787944025836, 0, 0, 0, 0 ]

[ "This invention relates, in general, to female urination devices, and, in particular, to female urination devices which allow a female to urinate in a standing position." ]

{ "pile_set_name": "USPTO Backgrounds" }

[ 0 ]

[ "Analysis and modeling of heat-labile enterotoxins of Escherichia coli suggests a novel space with insights into receptor preference.", "\nFeatures of heat-labile enterotoxins of Escherichia coli which make them fit to use as novel receptors for antidiarrheals are not completely explored. ", "Data-set of 14 different serovars of enterotoxigenic Escherichia coli producing heat-labile toxins were taken from NCBI Genbank database and used in the study. ", "Sequence analysis showed mutations in different subunits and also at their interface residues. ", "As these toxins lack crystallography structures, homology modeling using Modeller 9.11 led to the structural approximation for the E. coli producing heat-labile toxins. ", "Interaction of modeled toxin subunits with proanthocyanidin, an antidiarrheal showed several strong hydrogen bonding interactions at the cost of minimized energy. ", "The hits were subsequently characterized by molecular dynamics simulation studies to monitor their binding stabilities. ", "This study looks into novel space where the ligand can choose the receptor preference not as a whole but as an individual subunit. ", "Mutation at interface residues and interaction among subunits along with the binding of ligand to individual subunits would help to design a non-toxic labile toxin and also to improve the therapeutics." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.007575757575757576, 0.006578947368421052, 0.0125, 0, 0, 0, 0, 0, 0 ]

[ "This blog was written by Kevin Folta, chair of the horticultural department at the University of Florida. ", "Kevin hosts a Talking Biotech podcast each week, which the GLP posts each Monday. ", "This week’s podcast focused on the success of GMO brinjal (eggplant) in Bengladesh.", "\n\nBt brinjal is a transgenic eggplant with a natural protein gene inserted that naturally repels or kills many pests. ", "The protein is found in the soil bacterium Bacillus thuringiensis, which has been sprayed on organic crops for more than a century because of its effectiveness and harmlessness to humans and other crops. ", "Bt brinjal is genetically inoculated to resist the endemic pest called fruit and shoot borer, and therefore requires drastically less pesticides than is conventionally applied by farmers. ", "This is South Asia’s first GMO food crop, and has been developed in the public sector for distribution to resource-poor smallholder farmers.", "\n\n***\n\nIn general, the anti-GMO world has left me alone. ", "I don’t find much negativity about [my] podcast online . . . . ", "So imagine my surprise when I posted this week’s podcast on the Bt Brinjal (eggplant, aubergine). ", "The response was angry and vitriolic. . .", "\n\n\n\n\n\nWhat is happening here is amazingly obvious. . . . ", "these folks are really angry because the Bt Brinjal [which was developed in part by the Bengladesh government and whose seeds are sold to mostly poor farmers at a controlled price let’s all of the air out of their balloon]. ", "Free seeds, poor farmers, less insecticides, better products…. ", "where’s the downside?", "\n\n. . . [", "I]f this story becomes widely told, it represents the end of their war on science. ", "It is an example of how the technology should be used– for the betterment of human health and the environment.", "\n\nThey are having a remarkably hard time arguing with that.", "\n\nSo what happens when ideologically bent opinion can’t argue based on evidence? ", "It becomes an ad hominem effort to trash the messenger. . . .", "\n\n. . . . [", "W]hen the technology they oppose serves others, they are caught between acknowledging that it is doing good and abandoning their sacred belief that this technology can do no good– ever.", "\n\nThe GLP aggregated and excerpted this blog/article to reflect the diversity of news, opinion and analysis. ", "Read full, original post: Eggplant Podcast Sparks Angry Response" ]

{ "pile_set_name": "OpenWebText2" }

[ 0.018867924528301886, 0.036585365853658534, 0, 0, 0, 0, 0, 0, 0, 0.01020408163265306, 0, 0, 0.004464285714285714, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.009174311926605505, 0 ]

[ "Q:\n\nGenerics - call a method on every object in a List\n\nIs there a way to call a method on every object in a List - e.g.\nInstead of\nList<MyClass> items = setup();\n\nforeach (MyClass item in items)\n item.", "SomeMethod();\n\nYou could do something like\nforeach(items).SomeMethod();\n\nAnything built in, extension methods, or am I just being too damn lazy?", "\n\nA:\n\nYes, on List<T>, there is:\nitems.", "ForEach(item => item.", "SomeMethod())\n\nThe oddity is that this is only available on List<T>, not IList<T> or IEnumerable<T>\nTo fix this, I like to add the following extension method:\npublic static class IEnumerableExtensions\n{\n public static void ForEach<T>(this IEnumerable<T> items, Action<T> action)\n {\n foreach(var item in items)\n action(item);\n }\n}\n\nThen, you can use it on ANYTHING that implements IEnumerable<T>... not just List<T>\n\nA:\n\nitems.", "ForEach(i => i.SomeMethod());\n\nA:\n\nLists have a ForEach method. ", " You could also create an IEnumerable extension method, but it kind of goes against the spirit of Linq.", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0.006944444444444444, 0, 0, 0.008830022075055188, 0.015625, 0.009708737864077669, 0 ]

[ "Energy transfer of highly vibrationally excited azulene: collisions between azulene and krypton.", "\nThe energy-transfer dynamics between highly vibrationally excited azulene molecules and Kr atoms in a series of collision energies (i.e., relative translational energies 170, 410, and 780 cm(-1)) was studied using a crossed-beam apparatus along with time-sliced velocity map ion imaging techniques. \"", "Hot\" azulene (4.66 eV internal energy) was formed via the rapid internal conversion of azulene initially excited to the S4 state by 266-nm photons. ", "The shapes of the collisional energy-transfer probability density functions were measured directly from the scattering results of highly vibrationally excited or hot azulene. ", "At low enough collision energies an azulene-Kr complex was observed, resulting from small amounts of translational to vibrational-rotational (T-VR) energy transfer. ", "T-VR energy transfer was found to be quite efficient. ", "In some instances, nearly all of the translational energy is transferred to vibrational-rotational energy. ", "On the other hand, only a small fraction of vibrational energy is converted to translational energy (V-T). ", "The shapes of V-T energy-transfer probability density functions were best fit by multiexponential functions. ", "We find that substantial amounts of energy are transferred in the backward scattering direction due to supercollisions at high collision energies. ", "The probability for supercollisions, defined arbitrarily as the scattered azulene in the region 160 degrees <theta<180 degrees and DeltaEd>2000 cm(-1) is 1% and 0.3% of all other collisions at collision energies 410 and 780 cm(-1), respectively." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.010416666666666666, 0.0033222591362126247, 0.006756756756756757, 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "CPS to restrict game attendance\n\nLeague's away games barred to fans in a bid to cap violence\n\nJanuary 24, 2009|By Carlos Sadovi and Bob Sakamoto, TRIBUNE REPORTERS\n\nFaced with a recent rash of shootings and fistfights spurred by high school basketball games that left six students shot and others injured, Chicago Public Schools officials Friday cracked down on unruly crowds by restricting who can attend games.", "\n\nFans won't be allowed to attend away games, parents will have to get permission to watch their teens play at other schools, and it's possible no one will get to see a few particularly heated rivalries. ", "Starting times for varsity games also will be moved up to 4 p.m. The changes, which take effect Monday, apply to boys and girls games.", "\n\nThe district's decision upset some parents, including Ericka Davis. ", "Her son, Jermaine Winfield of North Lawndale College Prep, was shot Tuesday during an altercation following a game against ACT Charter at Collins High School.", "\n\n\"I'm all for a child's safety, but I can't agree with that [policy]. ", "I want to see my son's games. ", "How can they keep me away from watching my son play?\" ", "said Ericka Davis, adding she doesn't think she should have to apply for permission to go. \"", "The big issue here is the lack of security.\"", "\n\nIf anyone understands the changes, it's Winfield, who still has the bullet from a small-caliber handgun lodged in his thigh. ", "Winfield, 16, was on the bench Thursday to watch his teammates defeat host Marshall. ", "Moments after the victory, a brawl between two female fans broke out in the stands, and police responded.", "\n\n\"It's going to be harder to win a game at someone else's home court. ", "The home crowd gets the home team's players pumped up,\" Winfield said. \"", "If we have to play with no fans, that's going to be weird.\"", "\n\nWhile the district's new policies represent perhaps its strictest crackdown, authorities have been dealing with the issue of violence at games off and on for more than 40 years.", "\n\nDuring the civil unrest and racial strife of the 1960s, varsity basketball games started at 3:15 p.m., said Calvin Davis, the district's director of sports administration. ", "That lasted through the mid-1980s.", "\n\nIn early 2003, then-district CEO Arne Duncan, who recently left to become U.S. education secretary under President Barack Obama, encouraged principals to limit attendance to students and their relatives following a fight in the bleachers at Crane High School. ", "At the time, schools were hiring private security for games following a Chicago Police Department money-saving decision to stop assigning officers to games.", "\n\nWhile district officials say that its crowd-control policy has been on the books for at least 12 years, it was not enforced, Calvin Davis said. ", "Principals often looked the other way when visiting fans showed up, but under the latest crackdown the district is calling on principals to stop the practice.", "\n\nThe crackdown covers the rest of the regular season, which ends Monday, and the first two Public League playoff rounds, Davis said. ", "The final three playoff rounds will be played at local universities, which are exempt from the bans. ", "The plan calls for students to present their school identification cards to watch a game and parents who are visiting a school to watch their children will require permission from principals.", "\n\n\"I would prefer that no one from the visiting team attends the game, I believe the safety of our kids has to come first,\" Davis said.", "\n\nPeter Fosco, athletic director at Schurz High School, said earlier start times may present problems because the district does not offer bus service to get students to schools at the earlier times. ", "Varsity games currently begin about 6 p.m. with freshman and sophomore squads starting earlier. ", "Like most parents interviewed, Fosco said security at games should be beefed up.", "\n\n\"It's sad to say but we definitely need help from the Chicago police to show a presence before and after games, that's how things are going right now,\" he said.", "\n\nDavis, who met with the Police Department's top brass Friday, said there are about six district security officers -- some of whom may be off-duty Chicago police -- assigned to \"high-alert\" games where tensions are running high.", "\n\n- - -\n\nRecent rash of violence\n\nJan. 9: Five people were wounded when gunmen opened fire from a truck outside Dunbar Vocational Career Academy during a game against Hope College Prep. ", "A sixth person was reportedly injured when she was trampled. ", "The contest, which was about to enter the third overtime period, immediately was canceled.", "\n\nTuesday: Jermaine Winfield, a junior basketball player at North Lawndale College Prep, was shot in the leg after an altercation following his team's 76-47 victory over ACT Charter at Collins High School -- Lawndale's home court. ", "Winfield, 16, was released from the hospital a day later with the bullet from a small-caliber handgun still lodged in his thigh.", "\n\nTuesday: Simeon's 51-38 victory at Bogan was tainted when a fight broke out in the stands. ", "The game was called with 5 minutes, 55 seconds remaining in the fourth quarter. ", "Chicago police and security personnel restored order quickly, and no injuries were reported.", "\n\nThursday: With Winfield watching from the bench, North Lawndale defeated Marshall 66-63. ", "But a post-game fight between two female students turned into a bigger disturbance, with a police call for assistance resulting in eight squad cars and two wagons dispatched to Marshall. ", "Order was restored, and no injuries were reported." ]

{ "pile_set_name": "Pile-CC" }

[ 0.009708737864077669, 0, 0, 0.014285714285714285, 0.0189873417721519, 0, 0, 0, 0.010869565217391304, 0, 0.007874015748031496, 0.023529411764705882, 0, 0, 0.013888888888888888, 0, 0, 0.005747126436781609, 0, 0.011450381679389313, 0.00641025641025641, 0.00684931506849315, 0, 0.014925373134328358, 0, 0, 0.007407407407407408, 0.010050251256281407, 0, 0.0125, 0, 0.008733624454148471, 0.010752688172043012, 0, 0, 0.012987012987012988, 0.0078125, 0.010752688172043012, 0, 0, 0.03296703296703297, 0.0053475935828877, 0 ]

[ "Feasibility of reduction in use of the mouse bioassay: presence/absence screening for saxitoxin in frozen acidified mussel and oyster extracts from the coast of California with in vitro methods.", "\nIn the United States, the detection of paralytic shellfish poisoning (PSP) for regulatory purposes relies on the mouse bioassay (MBA). ", "Using a saxitoxin presence/absence test could reduce animal usage significantly. ", "Three in vitro methods, the RIDASCREEN Saxitoxin kit, MIST Alert, and a 5 h neuroblastoma assay, were evaluated in parallel with the MBA using 106 twice-frozen, acidified extracts from California-grown mussel and oyster tissues. ", "For each assay, a cutoff point was established whereby data below or equal to that point were scored as negative and were assigned a score of zero. ", "Data above the cutoff were considered positive and assigned a score of one. ", "Pearson correlation coefficients were generated. ", "The RIDASCREEN, MIST Alert, and neuroblastoma bioassay correlated to the MBA at 0.849, 0.853, and 0.832 when used for presence/absence detection. ", "These data suggest that a reduction in MBA usage could be achieved in the surveillance of California-grown mussels and oysters for PSP-associated toxins. ", "Correlation data between the in vitro assays, cost comparisons, and the potential for false negatives and false positives were examined. ", "Implications of these methodologies in protecting public health are discussed." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0.008733624454148471, 0, 0, 0, 0.00684931506849315, 0, 0, 0 ]

[ "Figures\n\nAbstract\n\nCaMdr1p is a multidrug MFS transporter of pathogenic Candida albicans. ", "An over-expression of the gene encoding this protein is linked to clinically encountered azole resistance. ", "In-depth knowledge of the structure and function of CaMdr1p is necessary for an effective design of modulators or inhibitors of this efflux transporter. ", "Towards this goal, in this study, we have employed a membrane environment based computational approach to predict the functionally critical residues of CaMdr1p. ", "For this, information theoretic scores which are variants of Relative Entropy (Modified Relative Entropy REM) were calculated from Multiple Sequence Alignment (MSA) by separately considering distinct physico-chemical properties of transmembrane (TM) and inter-TM regions. ", "The residues of CaMdr1p with high REM which were predicted to be significantly important were subjected to site-directed mutational analysis. ", "Interestingly, heterologous host Saccharomyces cerevisiae, over-expressing these mutant variants of CaMdr1p wherein these high REM residues were replaced by either alanine or leucine, demonstrated increased susceptibility to tested drugs. ", "The hypersensitivity to drugs was supported by abrogated substrate efflux mediated by mutant variant proteins and was not attributed to their poor expression or surface localization. ", "Additionally, by employing a distance plot from a 3D deduced model of CaMdr1p, we could also predict the role of these functionally critical residues in maintaining apparent inter-helical interactions to provide the desired fold for the proper functioning of CaMdr1p. ", "Residues predicted to be critical for function across the family were also found to be vital from other previously published studies, implying its wider application to other membrane protein families.", "\n\nAuthor Summary\n\nMembrane proteins belonging to the Major Facilitator Superfamily (MFS) transport molecules, including drugs, across the membrane and are known to be associated with drug resistance. ", "CaMdr1p is one such MFS major multidrug efflux pump whose over-expression is linked to frequently encountered azole resistance in hospital isolates of C. albicans. ", "Amino acid residues critical for a protein's function are conserved across members of the protein family. ", "However, the traditional measure of conservation is not a useful parameter in mapping a functionally important residue in membrane proteins e.g., hydrophobically conserved stretches form helical transmembrane regions of the protein and are responsible for membrane localization, which individually have limited effect on binding and transport. ", "We have developed a method that uses information theory to score the conservation of a residue relative to its context within the membrane and hypothesize that these residues would be critical for the protein's function. ", "The relevance of predicted residues in the functioning of MFS is validated on CaMdr1p.", "\n\nFunding: This work was supported in parts by grants to one of us (RP) from the Department of biotechnology (BT/PR9100/Med/29/03/2007, BT/PR9563/BRB/10/567/2009); Council of Scientific and Industrial Research (38 (1122)/06/EMR-II, 22/3/2006) and Department of Science and Technology (SR/SO/BB-12/2004, 5/9/2005). ", "KK acknowledges Council of Scientific and Industrial Research, India, for Senior Research fellowship. ", "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.", "\n\nCompeting interests: The authors have declared that no competing interests exist.", "\n\nIntroduction\n\nIn yeasts, including the pathogenic Candida, an up-regulation of multidrug transporter genes belonging to either ATP Binding Cassette (ABC) or Major Facilitator Superfamily (MFS) is frequently observed in the cells exposed to the drugs leading to the phenomena of multidrug resistance (MDR) [1]. ", "Among the 28 putative ABC and 95 MFS transporter genes identified in the C. albicans genome, only ABC transporters CaCdr1p and CaCdr2p and MFS transporter CaMdr1p, are found to be the major determinants of azole resistance [2],[3]. ", "The reversal of the functionality of these multidrug efflux pump proteins represents an attractive strategy to combat azole resistance.", "\n\nThe major ABC transporters such as CaCdr1p, CaCdr2p bear similar topology and exist as two homologous halves. ", "These, like any other member of the ABC superfamily have four distinct modules: two transmembrane domains (TMDs) each consisting of six transmembrane segments (TMSs) and two nucleotide binding domains (NBDs) located on the cytosolic side of the membrane. ", "Though similar in topology and promiscuity towards substrate specificity, these ABC multidrug transporters of C. albicans also display selectivity to the range of substrates they can export [4].", "\n\nThe transporters belonging to MFS, consists of membrane proteins from bacteria to higher eukaryotes and these are involved in symport, antiport or uniport of various substrates [5],[6]. ", "One of the 17 families of MFS transporters uses the proton motive force to drive drug transport and has been identified in both prokaryotes and eukaryotes [7]. ", "Crystal structures of MFS proteins such as lactose permease (LacY), glycerol-3-phosphate (GlpT), EmrD and oxalate: formate antiporter (OxlT), suggest high structural resemblance among this family of proteins [8]. ", "These consist of 12 TMS, arranged with a similar predicted topology, strongly supporting a common structural architecture or fold across all the MFS transporters [9]–[12]. ", "The fungal MFS members particularly those involved in drug transport are poorly explored in terms of their structure and function [13]. ", "The multidrug MFS transporter CaMdr1p belongs to DHA1 family which is widely distributed and includes both drug-specific and multidrug efflux pumps [14].", "\n\nRandom and site-directed mutational strategies have been extensively used to understand the structure and function of these MDR efflux proteins. ", "For example, random mutational analysis of an ABC transporter, ScPdr5p of budding yeast identified several amino acid residues that alter its substrate specificity and sensitivity to various inhibitors [15],[16]. ", "Tutulan-Cunita et al. ", "observed that several point mutations led to significant changes in drug specificity of ScPdr5p which are distributed throughout the length of the protein [17]. ", "Site-directed mutagenesis followed by an elegant screen done by Golin's group has revealed interactions between TMS 2 and the NBD which may help to define at least part of the translocation pathway for coupling ATP hydrolysis to drug transport mediated by ScPdr5p. ", "Recently, Schmitt et al. ", "have elucidated the role of H1068 in H-loop of ScPdr5p which couples ATP hydrolysis with drug transport [18].", "\n\nSite-directed mutational analysis of multidrug ABC multidrug transporter CaCdr1p (a close homologue of ScPdr5p) has revealed insight into its drug binding and efflux properties. ", "These studies have implicated some of the amino acid residues of TMS 5, 6, 11 and 12 as the components of the substrate binding pocket(s) of CaCdr1p [19],[20]. ", "Together, these studies suggest that the drug binding sites in CaCdr1p are scattered throughout the protein and probably more than one residue of different helices are involved in binding and extrusion of drugs. ", "However, there is still insufficient information available to predict where and how exactly the most common antifungals such as azoles bind and how are they extruded by CaCdr1p.", "\n\nSite-directed mutational strategies rely on conservation of residues in a Multiple Sequence Alignment (MSA). ", "The conservation of a residue is calculated from the amino acid frequency distribution in the corresponding column of a MSA. ", "However, the physicochemical conservation is not necessarily responsible for a protein's structure and function but could reflect a more general function such as membrane localization. ", "Thus conservation alone is not sufficient to distinguish between residues responsible for the protein function and membrane localization. ", "Membrane proteins differ from soluble proteins because of their inter-TM hydrophilic and TM hydrophobic propensities, which have allowed the development of efficient membrane protein TM prediction methods [21] and of membrane protein specific substitution matrices [22].", "\n\nThe quantification of residue conservation has evolved over the last few years to the use of information theoretic measures [23]. ", "Relative entropy is a distance measure commonly applied to multiple alignments by comparing the observed frequency distribution with a background distribution. ", "In the present study, we have developed and employed a new method using information theory to rationalize mutation strategies and also applied it to a MFS multidrug transporter CaMdr1p [24]. ", "Relative Entropy (RE) or the Kullback-Liebler divergence is an information theoretic measure of the difference between two probability distributions and has been increasingly applied in bioinformatics to identify functional residues [24],[25]. ", "The use of RE with background frequencies [26] can improve the prediction of a protein's functional residues [27]–[32] as well as detect residues that determine the functional subtype of proteins [28]. ", "Though the basic Kullback-Liebler equation has not changed, its intelligent application in our method calculates Relative Entropy (REM) relative to its context within the membrane. ", "The REM scoring scheme has been improved by treating TM and inter-TM regions of MFS proteins separately which has drastically increased the credibility over the existing methods [23]. ", "In this manuscript, we have compared traditional treatment of conservation, and standard RE, with our improved method. ", "We validated our predictions by replacing the predicted highest REM positions of CaMdr1p with alanine by site-directed mutagenesis. ", "We show that most of these residues when replaced with alanine showed decreased resistance to drugs which was corroborated by abrogated efflux of drugs. ", "Additionally, we could further confirm the functional relevance of each of these high REM residues by predicting their location in deduced 3D model of CaMdr1p and their role in maintaining apparent inter-helical interactions. ", "With this approach, our method enabled us to accurately predict MFS-wide function-specific residues, validated by using CaMdr1p.", "\n\nResults\n\nREM considers conservation as well as the background probability of each alignment position of a MSA\n\nA comprehensive non-redundant data set, sourced from all MFS sequences present in the 56.2 release, was generated. ", "This data set was then aligned using a membrane-specific multiple alignment program, which stacked the helices appropriately. ", "A highly conserved residue in a multiple alignment is predicted to have a functional significance. ", "We calculated conservation values using the algorithm from Jalview. ", "Residues shown to be conserved dominate the TM helices, and on closer evaluation are largely hydrophobic residues associated with membrane localization. ", "The traditional relative entropy and our modified treatment of the method (REM) were calculated on the same alignment using scripts written in-house. (", "Fig. ", "1 shows a representative section of the alignment along with the REM, RE and conservation scores; see supplementary Table S1 for the REM, RE and conservation scores for the entire MSA). ", "The distribution curve generated on the basis of REM from the MSA is shown in Fig. ", "2A. Notably the dominant signal using a conventional conservation measure lies in the TM helices and traditional RE also issues high scores to these residues which are less frequent in nature. ", "Our treatment using a REM dampens the membrane localization signals further increasing the signal from atypically occurring conserved residues (Figure S1). ", "Since REM considers conservation as well as the background probability of a residue at a particular alignment position, it is an improved index of the functional significance of a residue. ", "To emphasize this fact, thirty residues with highest values were short-listed each from the conservation list, RE list and REM list. ", "On comparison, it was found that there are thirteen residues shortlisted as both conserved and with high RE while it was found that only nine out of these short-listed columns are both conserved and have high REM (Supplementary Table S2). ", "The thirty alignment positions with high REM were further studied to assess their functional relevance. ", "Expectedly, all residues predicted using REM, are not present in every protein in the family. ", "In CaMdr1p, 16 residues were identical with the most frequently occurring residue in the thirty highest scored alignment columns, and were mutated to alanine to directly validate the prediction (Fig. ", "2B).", "\n\nFigure 1. ", "A portion of Multiple Alignment showing conservation and REM for each column.", "\n\nFigure showing a representative portion of the alignment of MFS sequences and is generated using Alscript [48]. ", "The alignment is coloured in a gradient from red to yellow on the basis of decreasing conservation score. ", "Conservation was calculated using a method by Livingstone et al.[49] and the scores are shown as a histogram. ", "The histogram compares the conservation, RE and the REM scores for the alignment columns shows in the figure. ", "Results of selected mutations in CaMdr1p are indicated by triangles, green being sensitive while red are resistant. ", "REM scores are better indicators of functional relevance than physicochemical conservation.", "\n\nPanel A: Histogram of the REM scores (+ve/background frequency) vs frequency for all positions of MSA of the MFS transporters. ", "The top 30 REM positions depicted in the boxed region were selected for further analysis. ", "The inset represents the same data by a line graph. ", "B: The table shows 16 out of the top 30 high REM alignment positions where the residue in CaMdr1p matched with the most frequent amino acid at that particular position in a MSA of 342 MFS members. ", "Their predicted location with respect to CaMdr1p is also displayed in the next column.", "\n\ndoi:10.1371/journal.pcbi.1000624.g002\n\nResidues with high REM are part of the known motifs of Major Facilitator Superfamily\n\nThese sixteen out of the top thirty positions, wherein the residue in CaMdr1p matched with the most occurring residue across that alignment position in MSA were analyzed for further studies by site-directed mutagenesis. ", "Interestingly, most of the sixteen residues with high REM turned out to be part of the well-known motifs of the MFS. ", "These motifs are identified as Motif A (GxLaDrxGrkxxl), Motif B (lxxxRxxqGxgaa) which are conserved throughout the MFS, Motif C (gxxxGPxxGGxl) only in 12 and 14-TMS family and Motif D2 exclusive to 12-TMS family [7]. ", "Three out of the sixteen residues short-listed for CaMdr1p; E178, G183 and R184 are a part of Motif A; residues L211, R215 and G219 are a part of Motif B and G256 is a part of motif C. In addition to the known motifs mentioned above, two new motifs have been identified by our study. ", "The residues in these stretches 273Wrxxf277 and 296Pespr300 have high REM scores. ", "However, the known motif D2 does not appear to be highly conserved in our alignment and is thus not predicted to be family-wide function-specific.", "\n\nSite-specific mutagenesis of residues with high REM shows that they are functionally critical\n\nAll the sixteen residues selected on the basis of high REM were mutated by employing site-directed mutagenesis and were replaced with alanine except G165, G183 and G256 which were replaced by leucine. ", "For functional analysis of the mutant variants, a heterologous hyper-expression system, where GFP-tagged CaMdr1p (CaMDR1-GFP) was stably over-expressed from a genomic PDR5 locus in a S. cerevisiae mutant AD1-8u−, was used [33]. ", "The host AD1-8u− developed by Goffeau's group, was derived from a Pdr1-3 mutant strain with a gain-of-function mutation in the transcription factor Pdr1p, resulting in constitutive hyper-induction of the PDR5 promoter [34]. ", "A single-copy integration of each transformant at the PDR5 locus was confirmed by Southern hybridization (data not shown). ", "Two positive clones of each mutant were selected to rule out clonal variations. ", "These residues with high REM score showed increased drug susceptibility and abrogated efflux of substrates such as [3H] MTX and [3H] FLU (Fig. ", "3). ", "Of note, there were few exceptions to the list of residues with high REM. ", "For example residues T160, L211, W273 and R274 have high REM values but do not appear to be critical for function of CaMdr1p since the drug susceptibility and efflux are not affected upon replacement of these residues with alanine.", "\n\nPanel A: Drug resistance profile of wild type and mutant CaMDR1-GFP yeast strains determined by spot assay. ", "For spot assay, cells were freshly streaked, grown overnight and then resuspended in normal saline to an A600 of 0.1 (1×106 cells) and 5 µl of five-fold serial dilutions, namely 1 (1:5), 2 (1:25), 3 (1:125) and 4 (1:625), of each strain was spotted on to YEPD plates in the absence (control) and presence of the following drugs: FLU (0.20 µg/ml), CYH (0.20 µg/ml), 4-NQO (0.20 µg/ml) and MTX (65 µg/ml). ", "Growth differences were recorded following incubation of the plates for 48 hrs at 30°C. ", "Growth was not affected by the presence of the solvents used for the drugs (data not shown). ", "B: [3H] MTX and [3H] FLU accumulation in the different mutant variants of CaMdr1p-GFP. ", "Controls AD1-8u− and RPCaMDR1-GFP have also been included for comparison. ", "The results are means±standard deviations for three independent experiments.", "\n\ndoi:10.1371/journal.pcbi.1000624.g003\n\nAll the mutant variants of CaMdr1p are properly surface localized\n\nTo confirm that the change in susceptibility observed in the mutant variants was not due to their poor expression or mislocalization, we compared the localization of GFP-tagged version of CaMdr1p (CaMDR1-GFP) and its mutant variants by FACS and confocal imaging. ", "A proper localization of all the mutant variants was confirmed by both these methods which showed proper rimmed appearance of GFP-tagged CaMdr1p. ", "The Western Blot analysis further confirmed that the expression levels of CaMdr1p of all the mutant variants were similar thus corroborating FACS and confocal data (Fig. ", "4).", "\n\nPanel A: Western Blot analysis of the PM fraction of mutant variants with anti-GFP antibody. ", "B: Confocal and FACS analysis of the all the mutant variants to check their expression and localization in comparison with AD1-8u− (negative control) and RPCaMDR1-GFP (positive control) [46].", "\n\nFor evaluating the relevance of high REM residues in CaMdr1p, a 3D homology model was constructed using available crystal structures of lac permease of E. coli (1pv6), glycerol-3-phophate of E. coli (1pw4) and oxalate: formate transporter of O. formigenes (1zc7) as described in Materials and Methods. ", "All the top thirty positions of highest REM are marked in the model which mostly lie in the N-terminal half of the protein (Fig. ", "5A). ", "Using the homology model, a symmetric contact map of CaMdr1p was generated as discussed in Materials and Methods (Fig. ", "5B). ", "We exploited this distance plot to ascertain the role of these residues with high REM.", "\n\nFigure 5. ", "The 3D homology model of CaMdr1p and the contact map derived from it.", "\n\nPanel A: The 3D homology model of CaMdr1p wherein the mutated residues are marked onto the model and are coloured on the basis of the phenotypes exhibited upon mutation. ", "Red denotes sensitive, green shows resistant and blue marks a position predicted to be important but not mutated as the CaMdr1p residues did not match the conserved residue in that alignment position. ", "The structure is viewed using Visual Molecular Dynamic (VMD) software. ", "B: The contact map of CaMdr1p is plotted between all the residues vs all the residues and displays the interactions between the beta carbon of each residue and beta carbon atoms of every other residue within and up to 8 A° of distance (Cα is used for glycine). ", "Each cross points to an interaction between a residue on x-axis and a residue on y-axis. ", "The lines represent the top thirty high REM residues and are coloured on the basis of the phenotypes where green represents the residues that are sensitive upon mutation while red are the ones that do not show any phenotype. ", "The blue lines mark the residues in CaMdr1p that did not match with the most frequent residue in that particular column of the MSA.", "\n\ndoi:10.1371/journal.pcbi.1000624.g005\n\nIt is apparent that residues L211, R215 and G219 in TMS 4 are within 8 A° distance to many residues of TMS 1, 2 and 3. ", "For example, it can be seen that residue G219 on TMS 4 lie on the same face of the helix and is within 8 A° to the residues G165 and G169 on TMS 2. ", "Indeed, the mutation of predicted G165 and G169 on TMS 2 resulted in abrogated drug susceptibility and transport (data not shown). ", "All the predicted interactions are summarized in Fig. ", "6A and shown in a pictorial representation of the homology model in Fig. ", "6B.\n\nPanel A: The table summarizes predicted inter-helical interactions mediated via selected residues with high REM. ", "More than one residue pair is predicted to be involved in maintaining the interactions between the helices. ", "B: Pictorial representation of inter-helical interactions via these high REM residues. ", "Figure shows that the residues involved in these interactions are majorly confined to the N-terminal half of the protein.", "\n\ndoi:10.1371/journal.pcbi.1000624.g006\n\nDiscussion\n\nThe multidrug MFS transporter CaMdr1p harbors a conserved antiporter ‘motif C’ within TMS 5. ", "Our recent study has revealed that the conserved and critical residues of this motif and of TMS 5 are bunched together on the same face of its helical wheel projection and are critical in drug efflux [35]. ", "However, the structure and function aspects of this major multidrug transporter remain poorly understood. ", "To address some of these questions, in this study, we have rationalized conventional mutational strategy and applied computational approach to predict functionally critical residues of CaMdr1p.", "\n\nThe sequence set described in this manuscript represents a comprehensive non-redundant coverage of annotated MFS sequences from SWISSPROT. ", "Many methods have been developed to improve the MSA of membrane protein families for accurate predictions of residues critical for structure and function [36]. ", "Membrane proteins have fold signals which are easily mapped to the primary sequence as TM and inter-TM stretches. ", "Considering the differences in physico-chemical properties of these two regions, membrane protein specific substitution matrices have been developed [22]. ", "However, we argued that a conservation score on the basis of identity or physico-chemical similarity still remains inadequate as the background frequencies of their immediate environmental milieu are radically different with respect to hydrophilic and hydrophobic propensities. ", "This is also apparent from the conservation scores of the MSA wherein a large proportion of the conserved columns correspond to hydrophobic TM regions. ", "Notably, two CaMdr1p residues (F216 and L217) with high conservation but low REM were taken as controls, when replaced with alanine showed no change in the phenotype (data not shown). ", "One of the most basic fold specific signals is the hydrophobic core in globular proteins, and the TM region in membrane proteins. ", "Unlike globular proteins, the hydrophobic TM region is continuous in the membrane protein's primary structure, and indeed this still remains one of the preferred methods to identify membrane proteins, and map their TM regions. ", "While it is intuitive that the synchronous stretch of hydrophobic residues is responsible for membrane localization, the application of a scoring method that can distinguish these residues from family-wide alignment columns associated with other functions has not yet been deployed. ", "In essence, we require a method that can objectively separate the TM signals from other signals. ", "To overcome these limitations, we improved existing method(s) of information theory wherein REM was calculated on the basis of MSA of MFS proteins, keeping in mind the differences in the environmental milieus. ", "We thus treated TM and inter-TM regions by different background probabilities for calculation of REM. ", "These REM scores helped us to predict those sites which have amino acid distributions very different from the respective background distribution thereby statistically predicted to be functionally critical. ", "Not all the residues predicted using REM, are present in every protein in the family. ", "In CaMdr1p, 16 residues were identical with the most frequently occurring residue in the thirty highest scored alignment columns, and were mutated to directly validate the prediction (Fig. ", "2B). ", "Our results of drug susceptibility assays revealed that almost all of these matched residues with high REM when replaced with alanine displayed sensitivity to the tested drugs and showed abrogated drug transport (Fig. ", "3). ", "Interestingly, when we mutated residues which had high conservation values but lower REM values (negative control); none showed alterations in drug susceptibilities and thus did not retain the functionally critical stringency as was evident from residues with higher REM. ", "For example, analysis of a few conserved columns of the MSA, such as F216, L217 and L171 having REM values between 0.57-0.44 revealed that their replacement with alanine did not affect the function of CaMdr1p (data not shown). ", "This strengthens the fact that our method takes into account the conservation along with the background frequency and thus lists out residues which affect the function. ", "Also, to check the efficiency of the method, another negative control used was to mutate residues which are having low conservation and low REM values but lie in the vicinity of one of the 16 selected high REM residues. ", "For example, when C225 which is closer to the critical G219 and D235, was mutated to C225A, the functioning of the protein was not affected (data not shown). ", "Similarly, for critical G256, when residues A248, A253 and V254 which are within its vicinity were mutated as A248G, A253G and V254A, the mutant variants continued to behave as WT-CaMDR1-GFP [35].", "\n\nTo further elucidate the role of predicted residues in the functionality of CaMdr1p, a homology model based on the available crystal structures of lac permease, glycerol-3-phophate and oxalate: formate transporter was deduced [9]-[11]. ", "The REM method predicts the relative importance of a residue purely from sequence analysis and is independent of the protein's structure. ", "However, the role a residue plays in the protein's function is not readily apparent from its sequence. ", "We exploited the protein's 3D model as a guide to reason why a residue is functionally critical. ", "The deduced 3D model suggested that similar to other MFS structures, the 12 TM helices of the CaMdr1p span the membrane in such a way that they form the channel pore particularly aligned by residues of TMS 2, 4, 5, 7, 8, 10 and 11. ", "From the deduced homology model of CaMdr1p, a symmetric contact map was generated to highlight the inter-helical interactions of the protein (Fig. ", "5B). ", "Based on the predictions from the distance map, we could show that many high REM residues are indeed a part of inter-helical interactions (Fig. ", "6B). ", "It is apparent that more than one residue pair is predicted to be involved in maintaining the interactions between helices (Fig. ", "6A).", "\n\nOur aim in developing this method was to identify residues with high specificity which would play a critical role across this entire MFS protein family. ", "Although signals associated with antiporter motifs have been identified using this method, a finer granularity in function such as substrate specificity determining residues is not expected, as these signals would not be family-wide. ", "Since the enlisted residues with high REM values which are functionally critical for CaMdr1p are expected to be family-wide function-specific and thus critical for the entire MFS protein data set, we validated their relevance from the earlier published work. ", "It is known that Motif A of the MFS transporters span an eight residue long loop between TMS 2 and 3 and is suggested to be involved in maintaining a β-turn linking the adjacent TM helices [14]. ", "In the present study, G183 and R184 in the loop between TMS 2 and TMS 3 of CaMdr1p were picked up as family-wide function-specific residues thus corroborating that these residues are a part of Motif A (GxLaDrxGrkxxl) which holds importance throughout the MFS transporters. ", "The hypothesized rocking motion in MFS presumably requires conformational changes in the TMS and the β–turns. ", "In this, the transporter inter-converts between Ci (inward facing) and Co (outward facing) states for translocation of substrates. ", "In glycerol-3-phosphate of E. coli, it was seen that D88 was involved in inter-conversion between these Ci and Co states of the protein [10]. ", "Interestingly, D88 corresponds to E178 of CaMdr1p which also lies in Motif A which upon mutation to alanine turns out to be critical for drug susceptibility and efflux (Table 1).", "\n\nTable 1. ", "High REM alignment positions shown to be critical for function in other MFS members.", "\n\ndoi:10.1371/journal.pcbi.1000624.t001\n\nMotif B (lxxxRxxqGxgaa) of all MFS has a role in energy coupling which spans the N-terminal half of TMS 4 [7]. ", "CaMdr1p contact map reveals that residues in Motif B interface with residues 165GxxxG169 on TMS 2. ", "Motifs rich in glycine and proline residues promote formation of special backbone conformation including kinks in TMS, tight interactions between TMS and very flexible β-turns. ", "In human VAchT, Motif B and the adjacent sequences contain a total of nine notch signatures. ", "A notch allows two helical TMS to approach each other unusually closely because small side chains are located at the interface. ", "R124 of PcaK of Pseudomonas putida which is equivalent to high REM R215 of CaMdr1p of C. albicans is shown to be critical for helix packing [37]. ", "Interestingly, G111 of LacY of E.coli which also occupies a position in the same alignment column is also critical and earlier shown to be a residue at a kink.", "\n\nResidues from Motif C (gxxxGPxxGGxl) which is exclusive to 12-TMS family are also picked up by our calculations [7]. ", "G150 of LacY of E.coli which is equivalent to high REM G256 of CaMdr1p is function-specific for LacY protein [38]. ", "A stretch of conserved residues 296Pespr300, previously unidentified, at the end of TMS 6 were also predicted with high REM. ", "We have mutated equivalent residues P296A, E297A and T298A of CaMdr1p that overlap with the consensus residues in the stretch and found that cells expressing these mutated variants displayed increased sensitivity to drugs. ", "However, the functional significance of these residues is yet to be established.", "\n\nThere are a few exceptions which emerged from our method. ", "For example, our method did not pick up any residue of Motif D2. ", "This could be an artifact of the method used for alignments in earlier studies. ", "In this study we have employed a membrane protein specific alignment method whereas earlier reports have used standard multiple alignments substitution matrices with smaller data sets. ", "However, when we repeated the alignment using MUSCLE [39] and with the standard substitution matrix (Blosum 62) [40] on the complete data set the motif still did not appear (data not shown). ", "Motif D2 is assumed to have a structural significance as it holds a major kink within TMS 1 but mutations in this motif do not alter the backbone conformation. ", "As an example of the possibly insignificant role of the motif, in human VAchT, the mutation of L49G in this motif completely eliminates propensity for a kink or notch and abolishes activity while normally a glycine itself is expected to be present at this position and is supposed to be involved in maintaining a major kink in this motif [41].", "\n\nOut of the 16 residues that were mutated, T160A, L211A, W273A and R274A did not lead to any phenotypic changes. ", "It is known that for some of the positions in alignment, the most frequent amino acid does not match with the residue of CaMdr1p at that site. ", "One reason for this could be that some of the functionally important residues co-evolve i.e., these residues may mutate, with compensatory mutation occurring elsewhere in the protein to regain function [42]. ", "T160 where the most frequent residue is a serine at that position may be one such case. ", "Another reason may be that the alignment used in this study involved prediction of TMS with the possibility of errors in demarcating the edges of TM helices. ", "Residues from columns lining the edges of the helices may be wrongly assigned to TM and inter-TM regions. ", "This probably explains the lack of any effect of mutation on residues T160 of TMS 2 and L211 of TMS 4 which lie at the edge of the respective TMS. ", "Other exceptions to our predictions are the mutation of W273 and R274 which though highly conserved and probably a part of the new motif but do not abrogate function upon mutation. ", "Although a few tryptophans in an ABC transporter MRP1, have been shown to be involved in substrate binding and transport [43], generally, in a membrane protein tryptophan residues located on the surface of the molecule are mainly positioned to form hydrogen bonds with the lipid head groups while their hydrophobic rings are immersed in the lipid part of the bilayer [44]. ", "We predict that W273 and R274 may be associated with membrane helix orientation and this function may not be perturbed by mutating them individually through alanine scanning. ", "Alternatively, the tryptophan-arginine residues could be functionally critical in tandem and compensate each other for the loss of either one of them.", "\n\nOf note, in our predictions, substrate specific residues with high REM are not picked up which predominantly occur in C-terminal of MFS proteins. ", "It should be mentioned that since our alignment considers the entire MFS, residues responsible for substrate specificity would only be selectively conserved within a subfamily and would not have sufficiently strong signals to be visible in this present family-wide study. ", "For this, the same method may be applied to a data set classified on the basis of substrate selectivity to identify residues critical to the functioning of that subfamily.", "\n\nThere are a number of conservation methods known but none has yet achieved both biological and statistical rigor. ", "We have used REM to separate conserved residues from the background function of TM localization. ", "The interpretations support the well-known fact that MFS has a conserved N-terminal half which has residues important for maintenance of a specific fold for this class of proteins while C-terminal half has a more specific role in substrate binding and recognition [7]. ", "Taken together, our study provides an insight into the molecular details of MFS transporters in general and CaMdr1p in particular. ", "Our method of scaled REM calculations improves its performance over other information theoretic methods. ", "Additionally, this study also provides a method for rational mutational analysis not only for MFS proteins but can be applied to any class of membrane proteins and thus makes it possible to predict and locate family-wide functionally relevant residues.", "\n\nMethods\n\nMultiple sequence alignment.", "\n\n561 MFS sequences having 12 TMS as predicted by TMHMM [21] were extracted from SWISSPROT release 56.2. ", "Redundancy in the sequences is reduced to 90% on the basis of identity using BLASTclust by setting the identity threshold S to 90 and keeping all other parameters to default values (http://genomes.ucsd.edu/manuals/blast/bl​astclust.html). ", "The resulting 342 sequences were then aligned by PRALINETM using TMHMM as the method for predicting TM lengths and keeping all other parameters at their default values [36],[45] (See Supplementary Dataset S1).", "\n\nCalculation of conservation score, RE and REM.", "\n\nConservation of an amino acid in an alignment column was calculated using Jalview using available scoring schemes which are restricted to soluble proteins, was modified to accommodate different background probabilities for TM and inter-TM regions.", "\n\nRE is calculated as the deviation of the amino acid distribution Pi(a) from a background disribution f(a).where Pi(a) is the probability of the occurrence of amino acid a in column i of the MSA f(a) is the background probability of amino acid a and is classically estimated from SWISSPROT as the probability of occurrence of an amino acid in a large data set of proteins.", "\n\nIn our method REM, f(a) is the background probability of amino acid a and is classically estimated as the probability of occurrence of an amino acid in a large data set of proteins. ", "f(a) = f(a)TM or f(a)iTM where f(a)TM and f(a)iTM is the background probability calculated separately for TM and inter-TM regions of the alignment. ", "The background probability is thus replaced with this environment specific background frequency. ", "A classical description of entropy also does not take into consideration the absence of data points which are gaps in the case of a multiple alignment column. ", "We used a scaling factor which is equal to number of amino acid positions excluding gaps in column i divided by total number of sequences. ", "The scores were then divided by this scaling factor to give scaled REM score.", "\n\nSite-directed mutagenesis of CaMdr1p.", "\n\nSite-directed mutagenesis was performed by using the Quick-Change Mutagenesis kit (Stratagene, La Jolla, CA, USA) as described previously [46]. ", "The mutations were introduced into the plasmid pRPCaMDR1-GFP according to the manufacturer's instructions, and the desired nucleotide sequence alterations were confirmed by DNA sequencing of the ORF. ", "The primers used for the purpose are listed in Supplementary Table S4. ", "The mutated plasmid, after linearising with Xba1, was used to transform AD1-8u− cells for uracil prototrophy by lithium acetate transformation protocol [46]. ", "Integration was confirmed by Southern Blot analysis (data not shown).", "\n\nPreparation of the plasma membranes and immunodetection of CaMdr1p and its mutant variants.", "\n\nThe plasma membranes (PM) were prepared from S. cerevisiae cells, as described previously [46]. ", "The PM protein concentration was determined by bicinchonic acid assay using bovine serum albumin as the standard. ", "For Western Blot analysis the immunoblot was incubated with anti-GFP monoclonal antibody (1:5000) (JL-8) (BD Biosciences) as described previously. ", "Immunoreactivity of GFP antibody was detected using goat anti-mouse horseradish peroxidase-labelled antibody (1:5,000) and was visualized using the enhanced chemiluminescence assay system (ECL kit, Amersham Biosciences, Arlington Heights, IL, USA) [46].", "\n\nDrug susceptibility.", "\n\nThe susceptibilities of yeast cells, harboring wild type CaMDR1-GFP and its mutant variants, were tested to different drugs by spot assay. ", "For spot assay, 5 µl samples of five-fold serial dilutions of yeast culture each with cells suspended in normal saline to an OD of 0.1 (1×106 cells) at A600 were spotted onto YEPD plates in the absence (control) or in the presence of the drugs [35]. ", "Growth differences were recorded following incubation of the plates for 48 hrs at 30°C.", "\n\nDrug transport of mutant variants.", "\n\nThe accumulation of [3H] MTX (specific activity, 8.60 Ci/mmol) and [3H] FLU (specific activity, 19 Ci/mmol) was determined by protocol described previously [35]. ", "Cells from mid-log phase were centrifuged at 500×g for 3 min and resuspended in fresh YEPD medium as 5% cell suspension. ", "100 µl of cell suspension was incubated in shaking water bath at 150 rpm at 30°C and [3H] MTX was added to a final concentration of. ", "The cells were incubated in [3H] MTX (25 µM) or [3H] FLU (100 nM) for 30 min, filtered rapidly and washed twice with 1x PBS, pH 7.4 on Millipore manifold filter assembly using 0.45 µm nitrocellulose filter discs (Millipore, U.S.A). ", "The filter discs were dried and put in cocktail-O and the radioactivity was measured in a liquid scintillation counter (Packard, Beckman, USA). ", "The accumulation was expressed relative to the wild type CaMdr1p-GFP.", "\n\nMolecular modeling of CaMdr1p.", "\n\nStructures of known MFS proteins [PDB id – 1pv6 [9], 1pw4 [10] and 1zc7 [11]] were retrieved from Protein Data Bank (www.rscb.org) and aligned using MODELLER9V5 [47]. ", "MFS sequences containing CaMdr1p aligned earlier with PRALINETM were aligned to this structure alignment after removing the sequences which corresponded to known structures. ", "These were aligned using the profile-profile alignment option of ClustalW. The Profile alignment was then manually refined based on helix packing information of known MFS structures. ", "Manual refinement was restricted to incorporating gaps in the loop regions without disturbing either the structural alignment or PRALINE sequence alignment. ", "The homology model for the target sequence CaMdr1p was generated using MODELLER9V5 [47] using 1pv6, 1pw4 and 1zc7 as template sequences. ", "The initial 90 residues from N-terminal of CaMdr1p did not align with any of the templates, and were omitted from further study. ", "The model was evaluated and validated by PROCHECK.", "\n\nGenerating the contact map.", "\n\nThe contact map displays the distances between Cβ of one residue and Cβ of every other residue within 8 A° distance (Cα is considered in case of glycine). ", "Using the homology model a symmetric contact map for CaMdr1p was generated using a PERL program written in-house.", "\n\nComparative plot of conservation (red), RE (green) and REM (blue) across the entire alignment. ", "The conservation scores are scaled for comparison with RE and REM. ", "Positions of the highest scoring alignment columns by each method are shown above the graph, along with the results of mutation of the matching residues. ", "Out of these top scoring positions by three different calculations, the mutated positions showing resistant phenotype are marked in red triangles, those showing sensitive on all drugs are marked in green triangles while those which were not mutated are marked by empty triangles. ", "Locations of the transmembrane regions are marked by black bars on the x-axis.", "\n\ndoi:10.1371/journal.pcbi.1000624.s006\n\n(0.03 MB PDF)\n\nAcknowledgments\n\nWe thank R. D. Cannon for providing us with the plasmid and the strains used in this study. ", "We further thank Ranbaxy Laboratories Limited, India for providing us with Fluconazole. ", "We are thankful to K. Natrajan and Devapriya Choudhury for their valuable comments on our manuscript." ]

{ "pile_set_name": "Pile-CC" }

[ 0.011111111111111112, 0, 0, 0, 0.007352941176470588, 0, 0, 0, 0, 0.005, 0.01, 0.006097560975609756, 0, 0, 0, 0.011627906976744186, 0.01592356687898089, 0.0196078431372549, 0, 0, 0.01282051282051282, 0.017241379310344827, 0, 0.017857142857142856, 0.00392156862745098, 0.005154639175257732, 0.005319148936170213, 0.00625, 0.009389671361502348, 0.011627906976744186, 0.007352941176470588, 0.006535947712418301, 0.006802721088435374, 0.004694835680751174, 0.045454545454545456, 0, 0.007547169811320755, 0.04, 0, 0.005555555555555556, 0, 0, 0, 0.018018018018018018, 0.008, 0.005405405405405406, 0, 0, 0, 0, 0.005235602094240838, 0.00819672131147541, 0, 0.011049723756906077, 0.010869565217391304, 0, 0.007575757575757576, 0, 0, 0.0078125, 0.008771929824561403, 0, 0, 0.014705882352941176, 0, 0, 0, 0.016129032258064516, 0.024096385542168676, 0, 0, 0.005291005291005291, 0.007518796992481203, 0, 0, 0.010638297872340425, 0.005, 0, 0, 0.025974025974025976, 0.008771929824561403, 0, 0, 0, 0, 0.01098901098901099, 0.015503875968992248, 0, 0, 0.01015228426395939, 0, 0.002881844380403458, 0.008547008547008548, 0.004608294930875576, 0, 0, 0, 0, 0.008771929824561403, 0, 0, 0, 0.006993006993006993, 0, 0.013513513513513514, 0.008658008658008658, 0, 0.0049504950495049506, 0, 0, 0.011494252873563218, 0.013513513513513514, 0, 0.0026954177897574125, 0.00684931506849315, 0.0058823529411764705, 0, 0, 0, 0, 0.007751937984496124, 0, 0.008403361344537815, 0, 0.011627906976744186, 0, 0, 0, 0, 0, 0.0038314176245210726, 0, 0, 0.007633587786259542, 0, 0, 0.007633587786259542, 0.018518518518518517, 0.0136986301369863, 0.00847457627118644, 0, 0, 0, 0.00684931506849315, 0, 0, 0, 0.014184397163120567, 0.00625, 0, 0, 0, 0.006578947368421052, 0.005434782608695652, 0, 0, 0, 0.010309278350515464, 0.009523809523809525, 0.0196078431372549, 0, 0.011627906976744186, 0.005291005291005291, 0, 0.0045871559633027525, 0, 0.007352941176470588, 0.022026431718061675, 0, 0.004545454545454545, 0, 0, 0, 0.007246376811594203, 0, 0, 0.008620689655172414, 0.006802721088435374, 0, 0.006944444444444444, 0, 0.007751937984496124, 0, 0.0064516129032258064, 0, 0.003861003861003861, 0.010256410256410256, 0.007326007326007326, 0.01818181818181818, 0.015267175572519083, 0.014084507042253521, 0.0056179775280898875, 0, 0.011904761904761904, 0.006578947368421052, 0, 0.011299435028248588, 0.010752688172043012, 0.0078125, 0, 0, 0, 0.008695652173913044, 0, 0.004484304932735426, 0, 0, 0, 0, 0, 0, 0, 0.0029154518950437317, 0, 0, 0, 0, 0.006329113924050633, 0, 0.006802721088435374, 0, 0.002680965147453083, 0, 0, 0.006756756756756757, 0.003676470588235294, 0, 0, 0, 0.0037174721189591076, 0.007633587786259542, 0, 0.003968253968253968, 0, 0.0380952380952381, 0.0041841004184100415, 0.004784688995215311, 0.020833333333333332, 0.004016064257028112, 0.002680965147453083, 0.005434782608695652, 0.006756756756756757, 0, 0, 0, 0, 0, 0.0136986301369863, 0, 0, 0, 0.014492753623188406, 0, 0.01020408163265306, 0, 0.013605442176870748, 0.007905138339920948, 0, 0, 0, 0, 0, 0.006097560975609756, 0, 0, 0.004310344827586207, 0.006944444444444444, 0, 0, 0.01775147928994083, 0.005747126436781609, 0.00546448087431694, 0, 0, 0, 0.02, 0, 0.006369426751592357, 0, 0.010309278350515464, 0.014925373134328358, 0, 0, 0, 0.012121212121212121, 0.022727272727272728, 0.009900990099009901 ]

[ "Pd-catalyzed decarboxylative cross coupling of potassium polyfluorobenzoates with aryl bromides, chlorides, and triflates.", "\nPd-catalyzed decarboxylative cross coupling of potassium polyfluorobenzoates with aryl bromides, chlorides, and triflates is achieved by using diglyme as the solvent. ", "The reaction is useful for synthesis of polyfluorobiaryls from readily accessible and nonvolatile polyfluorobenzoate salts. ", "Unlike the Cu-catalyzed decarboxylation cross coupling where oxidative addition is the rate-limiting step, in the Pd-catalyzed version decarboxylation is the rate-limiting step." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.00819672131147541, 0.005952380952380952, 0, 0 ]

[ "Vasopressin cells in the medial amygdala of the rat project to the lateral septum and ventral hippocampus.", "\nThe rat brain contains a large number of vasopressin (VP) immunoreactive fibers, the sites of origin of which have not yet been established completely. ", "For instance, the sources of VP fiber systems in the amygdala, ventral hippocampus (VH), mediodorsal thalamic nucleus, ventral tegmental area, and dorsal raphe yet remain obscure. ", "These VP fibers may originate in any of the recently described extrahypothalamic VP cell groups, viz., ", "medial amygdaloid nucleus (AME), dorsomedial hypothalamic nucleus, or locus coeruleus, since VP efferents from these cells still remain to be demonstrated. ", "In search of AME VP efferents three approaches were followed: (1) the Phaseolus vulgaris anterograde tracing method, (2) immunocytochemistry after AME lesioning, and (3) retrograde transport of a fluorescent dye in combination with immunofluorescence. ", "The results demonstrate that VP cells in the AME project to (1) the lateral septum (LS) by the ventral amygdalofugal pathway and (2) the VH via the amygdalohippocampal transition zone. ", "In addition, the VP projection from the bed nucleus of the stria terminalis (BST) to the LS was confirmed. ", "There was no indication that VP cells in the AME project through the amygdalotegmental pathway to the medulla oblongata and spinal cord. ", "The results support the possibility that the BST and AME are an anatomical entity that may be part of the central loci controlling sexual processes in the rat." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.009433962264150943, 0, 0.005555555555555556, 0.009708737864077669, 0.00641025641025641, 0, 0, 0.028037383177570093, 0, 0.006289308176100629 ]

[ "[Which is the preferred perioperative beta-blocker?].", "\nGuidelines on perioperative cardiovascular evaluation and management of patients undergoing non-cardiac surgery recommend initiation of beta-blocker therapy in at-risk patients who are undergoing intermediate- to high-risk surgery. ", "Continuation of therapy in patients already receiving beta-blockers is also recommended. ", "Recent literature, however, reported an increased risk of perioperative cardiovascular mortality among patients who continued with existing beta-blockade; most patients in this study were using metoprolol. ", "There are important pharmacodynamic and pharmacokinetic differences between various beta-blockers, and these differences may explain the differences in clinical effects. ", "Metoprolol has less beta1 receptor affinity compared with atenolol and bisoprolol, and beta1 receptor polymorphisms affect the clinical effects of metoprolol. ", "Furthermore, metoprolol is dependent on activity of the CYP2D6 liver enzyme, which results in clinically important differences in plasma concentration. ", "It is, therefore, wise to follow the European guidelines and to initiate beta-blocker therapy in the perioperative period with either atenolol or bisoprolol." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "Review of HPV-related diseases and cancers.", "\nHuman papillomavirus (HPV) is a double-stranded circular DNA virus belonging to the papillomavirus family. ", "It is transmitted by skin-to-skin or mucosa-to-mucosa contact and enters the body via cutaneous or mucosal trauma. ", "HPV infection is the most common sexually transmitted disease, although it is usually cured by the immune system. ", "Worldwide, the risk of being infected at least once in a lifetime among both men and women is 50%. ", "HPV infection causes common and anogenital warts, as well as other non-dermatological diseases. ", "The role of HPV in cancer development has been extensively studied, primarily in cervical cancer, but also in other types of neoplasms." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0.009259259259259259, 0, 0.008771929824561403, 0, 0.010416666666666666, 0.007407407407407408 ]

[ "Statutory pay\n\nPayments may be required from an employer if an employee is not at work for a variety of reasons.", "\n\nThere are detailed conditions for an employee to qualify for any of these statutory payments.", "\n\nEmployees are only eligible for a statutory payment if they have sufficient average weekly earnings of at least the lower earnings limit.", "\n\n2019/20:\n\nType\n\nMax payment period\n\nAmount (£)\n\nStatutory Sick Pay\n\n94.25\n\nStatutory Maternity Pay\n\nFirst six weeks\n\n90% of weekly earnings\n\nNext 33 weeks\n\n148.68\n\nStatutory Paternity Pay\n\n2 weeks\n\n148.68\n\nStatutory Adoption Pay\n\nFirst six weeks\n\n90% of weekly earnings\n\nNext 33 weeks\n\n148.68\n\nShared Parental Pay\n\n148.68\n\nAverage weekly earnings £118 or over.", "\n\nWith the exception of Statutory Sick Pay, statutory payments may be payable at 90% average weekly earnings throughout the payment period in certain circumstances. ", "This applies where 90% weekly earnings are less than the standard rate of £148.68.", "\n\n2018/19:\n\nType\n\nMax payment period\n\nAmount (£)\n\nStatutory Sick Pay\n\n92.05\n\nStatutory Maternity Pay\n\nFirst six weeks\n\n90% of weekly earnings\n\nNext 33 weeks\n\n145.18\n\nStatutory Paternity Pay\n\n2 weeks\n\n145.18\n\nStatutory Adoption Pay\n\nFirst six weeks\n\n90% of weekly earnings\n\nNext 33 weeks\n\n145.18\n\nShared Parental Pay\n\n145.18\n\nAverage weekly earnings £116 or over.", "\n\nWith the exception of Statutory Sick Pay, statutory payments may be payable at 90% average weekly earnings throughout the payment period in certain circumstances. ", "This applies where 90% weekly earnings are less than the standard rate of £145.18.", "\n\nStatutory Sick Pay\n\nPayments may be required from an employer if an employee is too ill to work.", "\n\nSSP is generally payable for a period up to 28 weeks.", "\n\nStatutory Maternity Pay\n\nPayments may be required from an employer when an employee takes time off to have a baby.", "\n\nSMP is payable for a period up to 39 weeks.", "\n\nStatutory Paternity Pay\n\nPayments may be required from an employer when an employee takes time off during their partner's Statutory Maternity Pay period.", "\n\nPayment is for a period of either one or two complete weeks.", "\n\nShared Parental Pay\n\nPayments may be required from an employer when an employee takes time off following the curtailment of the period of SMP by the mother.", "\n\nPayment is for up to a maximum of 37 weeks and is dependent on the mother's unused SMP period.", "\n\nStatutory Adoption Pay\n\nPayments may be required from an employer when an employee takes time off when they adopt a child." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0, 0, 0.0055248618784530384, 0.006060606060606061, 0, 0.0055248618784530384, 0.006060606060606061, 0, 0, 0, 0, 0, 0, 0, 0.006329113924050633, 0.010416666666666666, 0 ]

[ "As the 2020 census approaches, a timely exhibition at the Museum of the City of New York considers the history and meaning of America’s decennial head count.", "\n\nAround four years ago, when the Museum of the City of New York (MCNY) began considering an exhibition exploring the census, the 2020 decennial count had not yet become an explosive, headline-making political issue. ", "But for Joe Salvo, New York’s chief demographer, the importance of getting a full and fair count “was already weighing very heavily on his mind,” says Sarah Henry, the deputy director and chief curator at MCNY. ", "Henry’s conversations with Salvo about the census began while he was a consultant on MCNY’s immersive Future City Lab, a permanent exhibition deploying animated maps as a way to understand New York and its inhabitants.", "\n\nAs the 2020 count approached, “The need to talk about data and why the census matters became more urgent,” says Henry. ", "Accurate numbers—and a share of the billions in federal funding that comes with them—would be critical to New York’s future. ", "With so much at stake, “we thought about what role we can play as a museum to make this important act of civic participation compelling and interesting to the public,” Henry explains.", "\n\nThe resulting exhibition is Who We Are: Visualizing New York by the Numbers. ", "Supported by a $150,000 grant from The Andrew W. Mellon Foundation, Who We Are opens November 22 and runs through September 2020, encompassing the census counting period. ", "Designed by Brooklyn-based firm Isometric, the exhibition will feel like an immersive, multimedia installation. ", "Data visualizations, interactive digital experiences, and physical objects including works of art and historical documents will explain the origins and staggeringly complex mechanics of this constitutionally mandated population count and its attendant debates, controversies, and political uses.", "\n\nAmong the historic data visualizations on display are Lusk Committee maps from 1919, which used census figures to profile immigrant neighborhoods and inform law enforcement raids in areas allegedly populated by \"anti-American\" radicals.", "\n\nExamples of different types of census forms from previous decades will reveal the near-constant evolution of census questions, many of which continue to have implications for historically undercounted communities. “", "How we describe ourselves and what categories are available have changed over time and reflect the eras in which they were written,” says Henry. ", "Still, some historic material may feel relevant to the present moment. ", "Striking in this regard are the infamous Lusk Committee maps of 1919, which used census figures to isolate immigrant neighborhoods into “racial colonies.” ", "Based on this data, law enforcement conducted raids in areas thought to be populated by dangerous and “anti-American” radicals.", "\n\nPedro M. Cruz, John Wihbey, Avni Ghael, and Felipe Shibuya, \"Simulated Dendrochronology of U.S. Immigration, 1790–2016,\" 2019. ", "Courtesy of the artists.", "\n\nIn addition to archival material, the exhibition includes projects by contemporary artists and thinkers who are interested in both aesthetics and interpretation. ", "Data innovators Pedro M. Cruz, John Wihbey, Avni Ghael, and Felipe Shibuya enliven statistics with Simulated Dendrochronology of U.S. Immigration, 1790–2016, a series of unusually poetic infographics using tree rings as a visual metaphor for the history of immigration—each new arrival building on the structure of previous migrations to America. ", "Further, Giorgia Lupi, a data visualization specialist and partner in the New York design firm Pentagram, has designed a customized interactive experience especially for MCNY in which museum visitors answer a series of questions that are transposed into a visual census of all museumgoers combined.", "\n\nUltimately Who We Are is intended to raise awareness of how important a full census count is for New Yorkers. “", "How can you decide where you need more housing or transportation without the facts? ", "The numbers enable the city to plan for the future,” Henry says. “", "This is a foundational act of civic engagement. ", "You can’t count if you’re not counted.”", "\n\n^ Back to top" ]

{ "pile_set_name": "OpenWebText2" }

[ 0.006369426751592357, 0.004608294930875576, 0.014218009478672985, 0.0045871559633027525, 0.008264462809917356, 0, 0.00546448087431694, 0.012658227848101266, 0.005847953216374269, 0.008928571428571428, 0, 0.004201680672268907, 0, 0.006896551724137931, 0, 0.0064516129032258064, 0, 0.03875968992248062, 0, 0, 0.01440922190201729, 0.006711409395973154, 0, 0, 0.015151515151515152, 0, 0, 0 ]

[ "---\nabstract: 'Loss-based clustering methods, such as k-means and its variants, are standard tools for finding groups in data. ", "However, the lack of quantification of uncertainty in the estimated clusters is a disadvantage. ", "Model-based clustering based on mixture models provides an alternative, but such methods face computational problems and large sensitivity to the choice of kernel. ", "This article proposes a generalized Bayes framework that bridges between these two paradigms through the use of Gibbs posteriors. ", "In conducting Bayesian updating, the log likelihood is replaced by a loss function for clustering, leading to a rich family of clustering methods. ", "The Gibbs posterior represents a coherent updating of Bayesian beliefs without needing to specify a likelihood for the data, and can be used for characterizing uncertainty in clustering. ", "We consider losses based on Bregman divergence and pairwise similarities, and develop efficient deterministic algorithms for point estimation along with sampling algorithms for uncertainty quantification. ", "Several existing clustering algorithms, including k-means, can be interpreted as generalized Bayes estimators under our framework, and hence we provide a method of uncertainty quantification for these approaches.'", "\nauthor:\n- Tommaso Rigon\n- 'Amy H. Herring'\n- 'David B. Dunson'\nbibliography:\n- 'biblio.bib'\ntitle: A generalized Bayes framework for probabilistic clustering\n---\n\nIntroduction {#sec:intro}\n============\n\nCluster analysis is a canonical topic in statistical learning, which continues to draw substantial attention. ", "Particularly popular are “algorithmic” approaches such as k-means, its generalizations based on Bregman divergences [@Banerjee2005], and the k-medoids (<span style=\"font-variant:small-caps;\">pam</span>) algorithm; see e.g. @Kaufman1990, [@Hastie2008] and @Jain2010 for an overview. ", "These methods rely on reasonable heuristics typically aimed at minimizing a specific loss function, and they are widely employed because they are conceptually simple, computationally efficient (k-means), and/or robust to moderate perturbations of the data (<span style=\"font-variant:small-caps;\">pam</span>). ", "However, despite their popularity, these approaches lack uncertainty quantification. ", "As an example, consider the dataset on the left side of Figure \\[fig:intro\\]: although k-means provides a sensible solution, some observations lie at the cluster boundaries. ", "In order to quantify this phenomenon, one could consider the misclassification probability, namely the chance that an observation belongs to one of the other groups. ", "Unfortunately, it is currently unclear how to compute such probabilities for k-means.", "\n\nModel-based clustering based on mixtures provides a rigorous inferential foundation that naturally leads to probabilistic assessments [@Fraley2002; @Fruhwirth2019]. ", "While conceptually appealing, mixture models face substantial practical difficulties, including computational bottlenecks and sensitivity to misspecification. ", "Posterior inference for Bayesian mixture models is routinely conducted via Markov chain Monte Carlo (<span style=\"font-variant:small-caps;\">mcmc</span>) which, despite remarkable advances [e.g. @Stephens2000; @Jain2004; @Wang2011; @Zuanetti2019], remains computationally expensive and affected by the label-switching issue. ", "In addition, misspecified mixtures may have unreliable and overly complex clustering solutions, thus requiring robust procedures which may increase the computational burden; see e.g. @Rodriguez2014 [@Miller2019; @Lijoi2020] for some recent advances in the Bayesian setting.", "\n\n![", "Clustering and uncertainty quantification on a toy example with $n = 50$ observations. ", "Left plot: data points, with colors representing the k-means solution. ", "Right plot: generalized Bayes misclassification probabilities associated with k-means. ", "\\[fig:intro\\]](intro){width=\"\\textwidth\"}\n\nIn this paper we aim at bridging the loss- and model-based frameworks through a generalized Bayes approach that provides a probability distribution characterizing uncertainty in clustering without requiring one to specify a likelihood for the data. ", "We do this within a coherent Bayesian framework for updating prior beliefs using information in the data. ", "Let $\\bm{X}$ be the data, $\\bm{\\theta}$ a parameter of interest and $\\pi(\\bm{\\theta})$ the corresponding prior. ", "Then, one can consider the generalized posterior $$\\label{GibbsPosterior}\n\\pi(\\bm{\\theta} \\mid \\lambda, \\bm{X}) \\propto \\pi(\\bm{\\theta})\\exp\\{ - \\lambda \\ell(\\bm{\\theta}; \\bm{X}) \\},$$ with $\\lambda > 0$ and $\\ell(\\bm{\\theta}; \\bm{X}) > 0$ an arbitrary loss function. ", "Expression (\\[GibbsPosterior\\]) is known as the *Gibbs posterior* [@Shawe1997; @McAllester1998; @Jiang2008]. ", "Standard Bayesian inference is a special case of , occurring when the loss function $\\lambda \\ell(\\bm{\\theta}; \\bm{X})$ is a negative log-likelihood. ", "Although Gibbs posteriors have been used in a wide variety of settings, it is only recently that they were shown to provide a rational update of beliefs and hence can be regarded as genuine posterior distributions [@Bissiri2016]. ", "@Holmes2017 provide an important development in developing methods for choosing $\\lambda$.\n\nDespite the increasing theoretical support for equation , to our knowledge the Gibbs posterior framework has not yet been used to define a general methodology for clustering. ", "There is a small literature on generalized Bayes methods for clustering, but these approaches tend to lack the link with loss function based methods such as k-means. ", "Importantly, they cannot be used to address our interest in providing uncertainty quantification for k-means and related methods. ", "A key recent example is the coarsened posterior framework of @Miller2019, which they use to provide a robust approach for Bayesian model-based clustering. ", "Refer to @Gorsky2020 for a successful application of this framework. ", "Alternatively, @Duan2019 propose a Bayesian model for a pairwise distance matrix representation of the data. ", "Early ideas for clustering estimation that combine prior information with general losses appear in @Lau2007. ", "Also relevant are the algorithms of @Kulis2012 [@Broderick2013], which are based on the very different idea of starting with a Bayesian mixture model and taking a small variance limit to define algorithms for clustering.", "\n\nIn contrast, we propose a very broad Gibbs posterior-based framework for uncertainty quantification in clustering. ", "A rich class of loss functions for clustering are considered, and we extensively study the resulting properties. ", "We additionally introduce novel algorithms for point estimation and uncertainty quantification, including for k-means and other existing clustering algorithms. ", "Our framework is also appealing in terms of simplicity and interpretability of tuning parameters.", "\n\nThe paper is organized as follows. ", "In Section \\[sec:method\\] we propose a generalized Bayes product partition modeling (<span style=\"font-variant:small-caps;\">gb-ppm</span>) framework and discuss properties. ", "In Section \\[sec:posterior\\_inference\\] we propose algorithms for posterior inference. ", "In Section \\[sec:Bregman\\] we study the use of Bregman divergences within <span style=\"font-variant:small-caps;\">gb-ppm</span>, and develop highly efficient computational algorithms. ", "In Section \\[sec:distance\\] we consider pairwise dissimilarities, show robustness to contamination, and develop a novel k-dissimilarities algorithm for point estimation. ", "The above classes are extremely broad, and we discuss important special cases in Section \\[sec:models\\]. ", "In Section \\[sec:illustration\\] we illustrate the methods through simulation studies and a medical application. ", "Concluding remarks are given in Section \\[sec:discussion\\].", "\n\nGeneralized Bayes clustering {#sec:method}\n============================\n\nBackground and Motivation {#sec:background}\n-------------------------\n\nLet $\\bm{x}_i = (x_{i1},\\dots, x_{id})^\\intercal$ be a vector of observations on $\\mathds{X} \\subseteq \\mathds{R}^d$, for $i=1,\\dots,n$, and let $\\bm{X}$ be the collection of all the data points. ", "Moreover, suppose $\\bm{C} = \\{C_1,\\dots,C_K\\}$ is a partition of the integers $\\{1,\\dots,n\\}$ into $K = |\\bm{C}|$ disjoint sets. ", "Alternatively, the partition can be described through cluster labels $\\bm{c} = (c_1,\\dots,c_n)$, so that $c_i = k$ for $i=1,\\dots,n$ if and only if $i \\in C_k$. In this case, only the induced partition is relevant and the labels on the indicators $\\bm{c} = (c_1,\\dots,c_n)$ are not of interest. ", "Typical Bayesian models for clustering are based on posterior distributions of the form $$\\label{mixturemodel}\n\\pi(\\bm{c} \\mid \\bm{X}) \\propto \\pi(\\bm{c}) \\prod_{k=1}^K \\left[\\int_{\\Theta} \\prod_{i \\in C_k} \\pi(\\bm{x}_i \\mid \\bm{\\theta}) \\pi(\\bm{\\theta}) { \\,\\textup d}\\bm{\\theta}\\right],$$ where $\\pi(\\bm{c})$ is the prior probability of $\\bm{c}$, $\\pi(\\bm{x} \\mid \\bm{\\theta})$ is the within-cluster likelihood, and $\\pi(\\bm{\\theta})$ is the prior law generating the cluster-specific parameters.", "\n\nAlthough equation  forms the basis for a vast literature on Bayesian clustering, there are key practical problems that arise. ", "The integral is often not available in closed form, leading to computational complications. ", "More importantly, the posterior on clustering is very sensitive to the precise specification of $\\pi(\\bm{x} \\mid \\bm{\\theta})$ and $\\pi(\\bm{\\theta})$. The assumption is that observations are drawn from $$\\label{latent}\n(\\bm{x}_i \\mid \\bm{\\theta}_k, c_i = k) \\overset{\\textup{ind}}{\\sim} \\pi(\\bm{x}_i \\mid \\bm{\\theta}_k), \\qquad i \\in C_k, \\quad k=1,\\dots,K,$$ where $\\bm{\\theta}_k \\overset{\\textup{iid}}{\\sim} \\pi(\\bm{\\theta})$, for $k=1,\\dots,K$. There are two critical problems with (\\[latent\\]): (i) clustering may be a convenient simplification and there may not actually be distinct groups in the data, and (ii) even if there are, the distribution of the data within each cluster is unlikely to exactly follow the chosen distribution $\\pi(\\bm{x} \\mid \\bm{\\theta})$. Unfortunately, the posterior on clustering is highly sensitive to both issues, leading to a critical brittleness problem.", "\n\nGeneralized Bayes product partition models\n------------------------------------------\n\nWe introduce a class of *generalized Bayes product partition models (<span style=\"font-variant:small-caps;\">gb-ppm</span>)* for clustering. ", "This broad family is characterized by a factorized loss $\\ell(\\bm{c}; \\bm{X})$. We consider the case in which the number of clusters $K$ is fixed, e.g. because it is known or has been selected in an exploration phase. ", "We discuss the choice of $K$ in Section \\[sec:bissiri\\]. ", "Let $\\bm{X}_k = \\{\\bm{x}_i : i \\in C_k\\}$ denote the observations belonging to cluster $C_k$, for $k=1,\\dots,K$. Then, <span style=\"font-variant:small-caps;\">gb-ppm</span> models are characterized by losses admitting the factorization $$\\label{general_loss}\n\\ell(\\bm{c}; \\bm{X}) = \\sum_{k=1}^K\\sum_{i \\in C_k} \\mathcal{D}(\\bm{x}_i; \\bm{X}_k), \\qquad \\bm{c} : |\\bm{C}| = K,$$ where $\\mathcal{D}(\\bm{x}_i; \\bm{X}_k) \\ge 0$ is a function of $\\bm{x}_i$ and $\\bm{X}_k$ which quantifies the discrepancy of the $i$th unit from $k$th cluster. ", "As a default, we focus on uniform clustering priors having the form $$\\label{unifprior}\n\\pi(\\bm{c}) = \\frac{1}{\\mathcal{S}(n,K)}, \\qquad \\bm{c} : |\\bm{C}| = K,$$ where $\\mathcal{S}(n,K) = 1/K! ", "\\sum_{k=0}^K (-1)^{K-k}K!\\{(K-k)!k!\\}^{-1}k^n$ is the Stirling number of the second kind. ", "Prior  is uniform over partitions having $K$ components; our framework is easily modified to consider more elaborate clustering priors, but we focus on the uniform case throughout the paper.", "\n\n\\[product\\_partition\\] Let the loss function $\\ell(\\bm{c}; \\bm{X})$ be as in  and the prior $\\pi(\\bm{c})$ be as in . ", "The generalized Bayes posterior under a *generalized Bayes product partition model (<span style=\"font-variant:small-caps;\">gb-ppm</span>)* has the form $$\\pi(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\pi(\\bm{c})\\prod_{k=1}^K \\rho(C_k; \\lambda, \\bm{X}_k) \\propto \\prod_{k=1}^K \\exp\\left\\{ - \\lambda \\sum_{i \\in C_k} \\: \\mathcal{D}(\\bm{x}_i; \\bm{X}_k)\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K.$$\n\nThe quantity $\\pi(\\bm{c} \\mid \\lambda, \\bm{X})$ is a well-defined probability mass function, because the normalizing constant is such that $0 < \\sum_{\\bm{c} : |\\bm{C}| = K}\\pi(\\bm{c}) \\prod_{k=1}^K \\rho(C_k; \\lambda, \\bm{X}_k) < \\infty$. Our proposed <span style=\"font-variant:small-caps;\">gb-ppm</span> relates to the existing literature on Bayesian product partition models; refer to @Quintana2003 [@Park2010; @Muller2011] among others. ", "Such models use Definition \\[product\\_partition\\] to define a data-based prior for clustering, while inheriting the disadvantages described in Section \\[sec:background\\] in relying on (\\[mixturemodel\\]) and requiring specification of $\\pi(\\bm{x} \\mid \\bm{\\theta})$ and $\\pi(\\bm{\\theta})$.\n\nA primary contribution of this paper is to use the generalized posterior in Definition \\[product\\_partition\\] for inference on clustering including uncertainty quantification. ", "The <span style=\"font-variant:small-caps;\">gb-ppm</span> framework is extremely broad, and we will consider a variety of important subclasses corresponding to different choices of $\\mathcal{D}(\\bm{x}_i; \\bm{X}_k)$. Perhaps the simplest example is squared error loss as described in Example \\[kmeans\\].", "\n\n\\[kmeans\\] Let $||\\bm{x}||_2^2 = x_1^2 + \\cdots + x_d^2$ for any $\\bm{x} \\in \\mathds{X}$. Moreover, let $\\bar{\\bm{x}}_k$ represent the vector with the arithmetic means of the columns of $\\bm{X}_k$, for any $k=1,\\dots,K$. Then the loss $$\\label{kmeans_loss}\n\\ell(\\bm{c}; \\bm{X}) = \\sum_{k=1}^K \\sum_{i \\in C_k} ||\\bm{x}_i - \\bar{\\bm{x}}_k||_2^2,$$ defines a <span style=\"font-variant:small-caps;\">gb-ppm</span>.", "\n\nDecision theoretic justification {#sec:bissiri}\n--------------------------------\n\nWe now show that any <span style=\"font-variant:small-caps;\">gb-ppm</span> can be rightfully regarded as a posterior distribution, being a rational update of one’s prior beliefs, thus adapting the reasoning of @Bissiri2016 to the clustering setting. ", "In our <span style=\"font-variant:small-caps;\">gb-ppm</span> framework, the target of inference is an optimal and unknown partition $\\bm{c}_\\textsc{opt}$, defined as the minimizer of an integrated loss function, namely $$\\label{copt}\n\\bm{c}_\\textsc{opt} = \\arg \\min_{\\bm{c}} \\mathds{E}_{\\pi_0}\\{ \\ell(\\bm{c}; \\bm{X}) \\} = \\arg \\min_{\\bm{c} : |\\bm{C}| = K} \\sum_{k=1}^K\\sum_{i \\in C_k} \\mathds{E}_{\\pi_0}\\left\\{ \\mathcal{D}(\\bm{x}_i; \\bm{X}_k)\\right\\},$$ where the expectation is taken with respect to the unknown data generating process $\\pi_0(\\bm{X})$. The term $\\mathds{E}_{\\pi_0}\\{\\ell(\\bm{c}; \\bm{X}) \\}$ is sometimes called the *frequentist risk*. ", "The definition of $\\bm{c}_\\textsc{opt}$ does not require the existence of a true partition of the data. ", "Even in the absence of truly distinct groups of observations within a dataset, it is often of interest to infer clusters that capture aspects of the data encoded by the loss function.", "\n\nThe target $\\bm{c}_\\textsc{opt}$ cannot be estimated based on a single dataset without knowledge of $\\pi_0(\\bm{X})$. Instead we rely on the <span style=\"font-variant:small-caps;\">gb-ppm</span> posterior, which can be viewed as the “best” conditional distribution for quantifying our subjective beliefs about $\\bm{c}_\\textsc{opt}$. Let $\\mathscr{P}$ be the space of all conditional distributions given the data and let $\\mathscr{L}$ be a loss function defined in such a space. ", "For any $\\nu_1,\\nu_2 \\in \\mathscr{P}$, the loss $\\mathscr{L}$ describes whether $\\nu_1$ is a better candidate than $\\nu_2$ for representing ones’ posterior beliefs about $\\bm{c}_\\textsc{opt}$. Let $\\pi(\\bm{c} \\mid \\lambda, \\bm{X}) = \\arg \\min_{\\nu \\in \\mathscr{P}} \\mathscr{L}\\{\\nu(\\bm{c})\\}$. A reasonable candidate for $\\mathscr{L}$ is the loss $$\\label{loss_gibbs}\n \\mathscr{L}\\{\\nu(\\bm{c})\\} = \\lambda \\mathds{E}_\\nu\\left\\{\\ell(\\bm{c}; \\bm{X})\\right\\} + \\textsc{kl}\\{\\nu(\\bm{c}) \\mid\\mid \\pi(\\bm{c}) \\},$$ where the expectation is taken with respect to $\\nu(\\bm{c})$. It is easy to show that the unique minimizer of equation  is indeed the <span style=\"font-variant:small-caps;\">gb-ppm</span> of Definition \\[product\\_partition\\].", "\n\nThe loss $\\mathscr{L}$ in equation  balances two components: the discrepancy with respect to the observed data $\\mathds{E}_\\nu\\left\\{\\ell(\\bm{c}; \\bm{X})\\right\\}$, and the closeness to the prior $\\textsc{kl}\\{\\nu(\\bm{c}) \\mid\\mid \\pi(\\bm{c})\\}$. The parameter $\\lambda > 0$ controls the weight assigned to the former component. ", "To clarify the relationship between the minimization problem in  and the loss $\\mathscr{L}$ of equation , consider the following extreme cases. ", "As $\\lambda \n\\rightarrow 0$ the closeness to the data is not penalized and therefore one obtains $\\pi(\\bm{c} \\mid 0, \\bm{X}) = \\pi(\\bm{c})$. As $\\lambda \\rightarrow \\infty$ the effect of the prior is negligible and the Gibbs posterior collapses to a point mass $\\delta_{\\hat{\\bm{c}}_\\textsc{opt}}$, where $\\hat{\\bm{c}}_\\textsc{opt} = \\arg \\min_{\\bm{c}} \\ell(\\bm{c}; \\bm{X})$ is the so called *empirical risk minimizer*, i.e. the empirical counterpart of $\\bm{c}_\\textsc{opt}$. Hence, the <span style=\"font-variant:small-caps;\">gb-ppm</span> posterior combines the empirical version of $\\bm{c}_\\textsc{opt}$ with one’s prior beliefs.", "\n\nThe above discussion suggests that a <span style=\"font-variant:small-caps;\">gb-ppm</span> should be regarded as a standard Bayesian posterior, with a slight but crucial distinction: the target of inference is not anymore some latent partition as in equation , but the more general object $\\bm{c}_\\textsc{opt}$. Therefore, within this context, the number of clusters $K$ should be regarded as the resolution at which we want to partition the observations rather than some attribute of the data that we should estimate. ", "Hence, we believe that the number of clusters $K$ should be subjectively specified as part of the loss $\\ell(\\bm{c}; \\bm{X})$, rather than inferred from the data. ", "Alternatively, $K$ could be selected in an exploration phase using standard strategies such as the “elbow” rule. ", "Refer also to Section \\[sec:illustration\\] for practical examples.", "\n\nPosterior inference {#sec:posterior_inference}\n===================\n\nPoint estimation and Gibbs sampling\n-----------------------------------\n\nWe develop a set of strategies for posterior inference for <span style=\"font-variant:small-caps;\">gb-ppm</span>s. We find that implementation is typically easier and much more efficient compared with mixture models. ", "In particular, very efficient algorithms for point estimation are available. ", "For uncertainty quantification, we rely on Gibbs sampling strategies that are straightforward and are often characterized by good mixing.", "\n\nWe first consider point estimation. ", "Although several alternatives exist [e.g. @Medvedovic2004; @Lau2007; @Fritsch2009; @Wade2018], maximum a posteriori (<span style=\"font-variant:small-caps;\">map</span>) estimation provides a particularly natural and simple choice. ", "In particular, we let $$\\hat{\\bm{c}}_{\\textsc{map}} = \\arg \\max_{\\bm{c}} \\pi(\\bm{c} \\mid \\lambda, \\bm{X}).$$ In a <span style=\"font-variant:small-caps;\">gb-ppm</span> the $\\hat{\\bm{c}}_{\\textsc{map}}$ can be equivalently obtained by minimizing the loss function $\\ell(\\bm{c}; \\bm{X})$ over the space of partitions having $K$ components. ", "Hence, $\\hat{\\bm{c}}_\\textsc{map}$ coincides with the empirical risk minimizer $\\hat{\\bm{c}}_\\textsc{opt}$. The proof of the following Proposition is trivial, but the practical implications are important.", "\n\n\\[map\\] Let $\\pi(\\bm{c} \\mid \\lambda, \\bm{X})$ be a <span style=\"font-variant:small-caps;\">gb-ppm</span>. ", "Then, $$\\hat{\\bm{c}}_{\\textsc{map}} = \\arg \\min_{\\bm{c} \\: : \\: |\\bm{C}| = K} \\sum_{k=1}^K\\sum_{i \\in C_k} \\mathcal{D}(\\bm{x}_i; \\bm{X}_k).$$\n\nThe estimate $\\hat{\\bm{c}}_\\textsc{map}$ in Proposition \\[map\\] does not depend on $\\lambda$. It turns out that $\\hat{\\bm{c}}_{\\textsc{map}}$ can be efficiently computed in several cases, as it coincides with the solution of well known clustering methods. ", "This links the algorithmic approaches to our framework, as the following example clarifies.", "\n\nIn a <span style=\"font-variant:small-caps;\">gb-ppm</span> with loss function  the <span style=\"font-variant:small-caps;\">map</span> estimate is $$\\hat{\\bm{c}}_{\\textsc{map}} = \\arg \\min_{\\bm{c} \\: : \\: |\\bm{C}| = K} \\sum_{k=1}^K\\sum_{i \\in C_k} ||\\bm{x}_i - \\bar{\\bm{x}}_k||_2^2, \\qquad \\bm{c} : |\\bm{C}| = K,$$ which corresponds to the k-means estimator [e.g. @Jain2010].", "\n\nFor non-<span style=\"font-variant:small-caps;\">map</span> point estimation and uncertainty quantification, we instead rely on Gibbs sampling. ", "Let $\\bm{c}_{-i} = (c_1,\\dots,c_{i-1},c_{i+1},\\dots,c_n)$ be the collection of cluster indicators without the $i$th unit, and let $\\{C_{1,-i},\\dots,C_{K,-i}\\}$ be the associated partition, data points $\\bm{X}_{k,-i} = \\{\\bm{x}_i : i \\in C_{k,-i}\\}$, and cohesion functions $\\rho(C_{k,-i}; \\lambda, \\bm{X}_{k,-i})$. In Gibbs sampling we cyclically re-allocate the indicators $c_i$ by sampling from their full conditionals. ", "In a <span style=\"font-variant:small-caps;\">gb-ppm</span> we focus on partitions $\\bm{c}$ such that $|\\bm{C}| = K$, implying that if the $i$th unit is the only element in a cluster, it can not be re-allocated.", "\n\n\\[GibbsSampling\\] Let $\\pi(\\bm{c} \\mid \\lambda, \\bm{X})$ be a <span style=\"font-variant:small-caps;\">gb-ppm</span>. ", "Then, the conditional distribution of $c_i$ given $\\bm{c}_{-i}$ is $$\\mathds{P}(c_i =k \\mid \\bm{c}_{-i}, \\lambda, \\bm{X}) \\propto \\frac{\\rho(C_k; \\lambda, \\bm{X}_k)}{\\rho(C_{k,-i}; \\lambda, \\bm{X}_{k,-i})} = \\exp\\left\\{ - \\lambda \\left[\\sum_{i' \\in C_k} \\mathcal{D}(\\bm{x}_{i'}; \\bm{X}_k) - \\sum_{i' \\in C_{k,-i}} \\mathcal{D}(\\bm{x}_{i'}; \\bm{X}_{k,-i}) \\right] \\right\\},$$ for $k = 1,\\dots,K$ and for any partition $\\bm{c} : |\\bm{C}| = K$.\n\nThis stochastic allocation has an intuitive interpretation, since the involved probabilities are ratios of cohesion functions. ", "In other words, the $i$th unit is likely to be allocated in the $k$th cluster if the cohesion of the newly created cluster $\\rho(C_k; \\lambda, \\bm{X}_k)$ is higher than the old cohesion $\\rho(C_{k,-i}; \\lambda, \\bm{X}_{k,-i})$.\n\nGibbs sampling strategies based on the re-allocations of the labels are not novel [e.g. @Escobar1995], and they are generally affected by slow mixing e.g. when transitioning from $K$ to $K'$ clusters. ", "However, this issue does not occur in <span style=\"font-variant:small-caps;\">gb-ppm</span>s because the number of groups is fixed. ", "Indeed, we are considering a simplified sampling problem characterized by fewer local optima.", "\n\nClustering validation and uncertainty quantification {#sec:uq}\n----------------------------------------------------\n\nA cluster solution is especially useful when it can be qualitatively validated and interpreted. ", "To this purpose, we adapt the notion of *centroid* and *medoid* to our context. ", "Secondly, we emphasize that uncertainty quantification can be very helpful in practice, because it provides a more complete description of the clustering problem. ", "To illustrate this crucial aspect, we review several inferential quantities that one may want to consider in a cluster analysis.", "\n\nLet $\\bm{m}_k \\in \\mathds{X}$ denote the centroid of the $k$th cluster. ", "Each term $\\rho(C_k; \\lambda, \\bm{X}_k) = \\prod_{i \\in C_k} \\rho(\\bm{x}_i; \\lambda, \\bm{X}_k)$ can be written as the product of unit-specific contributions, each representing the cohesion of the $i$th unit to the $k$th cluster. ", "Hence, given a specific partition $C_1,\\dots,C_K$ and recalling that $\\bm{X}_k = \\{\\bm{x}_i : i \\in C_k\\}$, we define the associated centroids $\\bm{m}_1,\\dots,\\bm{m}_K$ as $$\\bm{m}_k = \\arg \\max_{\\bm{x} \\in \\mathds{X}} \\rho(\\bm{x}; \\lambda, \\bm{X}_k) = \\arg \\min_{\\bm{x} \\in \\mathds{X}}\\mathcal{D}(\\bm{x}; \\bm{X}_k), \\qquad k=1,\\dots,K,$$ that is, the centroid is defined as the value having the maximal cohesion within the $k$th cluster. ", "For instance, in the k-means Example \\[kmeans\\] one has that $\\bm{m}_k = \\arg \\min_{\\bm{x} \\in \\mathds{X}} ||\\bm{x} - \\bar{\\bm{x}}_k||_2^2$, implying that $\\bm{m}_k = \\bar{\\bm{x}}_k$, thus recovering the usual definition of centroid. ", "A closely related alternative is the so called medoid, which is a representative data point $\\bm{x}_{i^*_k}$ of the $k$th cluster. ", "The associated indices are obtained as $$i^*_k = \\arg \\max_{i \\in C_k} \\rho(\\bm{x}_i; \\lambda, \\bm{X}_k) = \\arg \\min_{i \\in C_k}\\mathcal{D}(\\bm{x}_i; \\bm{X}_k), \\qquad k=1,\\dots,K.$$ The medoids can be always computed because they involve a search over a finite set of points. ", "We will show in Section \\[sec:Bregman\\] that the centroids can be also easily obtained in most cases.", "\n\nWe now build on existing tools for uncertainty quantification in clustering problems. ", "In principle, it would be interesting to consider the probabilities $\\mathds{P}(c_i = k \\mid \\lambda, \\bm{X})$, but these are affected by the label-switching phenomenon [@Stephens2000]. ", "For this reason, one typically focuses on the so-called co-clustering matrix $\\bm{S}$ [e.g. @Fritsch2009], whose entries $s_{ii'}$ are such that $$s_{ii'} = \\mathds{P}(c_i = c_{i'} \\mid \\lambda, \\bm{X}), \\qquad i,i' \\in \\{1,\\dots,n\\}.$$ Each probability $s_{ii'}$ can be easily approximated from <span style=\"font-variant:small-caps;\">mcmc</span> samples and does not depend on the specific labels of $c_i$ and $c_{i'}$. From the matrix $\\bm{S}$, one can extract the probabilities $$\\label{miscl_probabilities}\ns_{ii^*_k} = \\mathds{P}(c_i = c_{i_k^*} \\mid \\lambda, \\bm{X}), \\qquad i = 1,\\dots,n, \\quad k=1,\\dots,K,$$ with $i_k^*$ being the $k$th medoid. ", "These probabilities can be viewed as surrogates for the unavailable $\\mathds{P}(c_i = k \\mid \\lambda, \\bm{X})$. The quantities $s_{ii^*_k}$ can be used to identify those units whose clustering allocation is uncertain, as in Figure \\[fig:intro\\]. ", "See Section \\[sec:illustration\\] for further practical examples.", "\n\nSuppose a new data point $\\bm{x}_{n+1}$ becomes available and we are interested in assigning it to a cluster. ", "The allocation probabilities are readily available from Theorem \\[GibbsSampling\\], and one has that $$\\label{cluster_predictive}\n\\mathds{P}(c_{n+1} = k \\mid \\lambda, \\bm{X}, \\bm{x}_{n+1}) \\propto \\frac{\\rho(C_{k,n+1}; \\lambda, \\bm{X}_{k,n+1})}{\\rho(C_k; \\lambda, \\bm{X}_k)}, \\qquad k=1,\\dots,K,$$ where $C_{k,n+1}$ and $\\bm{X}_{k,n+1}$ represent the $k$th index set and data matrix, respectively, having added the $n+1$th observation.", "\n\nA decision-theoretic approach for uncertainty quantification is described in @Wade2018, who propose point estimates and credible intervals for random partitions. ", "Their ideas rely on the *variation of information* (<span style=\"font-variant:small-caps;\">vi</span>) distance $d_\\textsc{vi}(\\bm{c},\\bm{c}')$ [@Meila2007]. ", "To characterize uncertainty around a point estimate $\\hat{\\bm{c}}$ they consider a credible ball $\\mathcal{B}_{1-\\alpha}(\\hat{\\bm{c}})$ having probability $1 - \\alpha$. Such a ball is constructed by adding partitions of increasing <span style=\"font-variant:small-caps;\">vi</span> distance from $\\hat{\\bm{c}}$ until the posterior probability is greater than $1-\\alpha$. To represent the boundaries of $\\mathcal{B}_{1-\\alpha}(\\hat{\\bm{c}})$, @Wade2018 use *horizontal bounds* $\\hat{\\bm{c}}_{1-\\alpha}$ corresponding to the partitions in $\\mathcal{B}_{1-\\alpha}(\\hat{\\bm{c}})$ having the greatest <span style=\"font-variant:small-caps;\">vi</span> distance from $\\hat{\\bm{c}}$. If $\\hat{\\bm{c}}$ and $\\hat{\\bm{c}}_{1-\\alpha}$ are similar, then posterior uncertainty in clustering is small.", "\n\nCohesions based on Bregman divergences {#sec:Bregman}\n======================================\n\nWe now specialize the general <span style=\"font-variant:small-caps;\">gb-ppm</span> formulation and describe a family of cohesion functions based on Bregman divergences. ", "The $\\hat{\\bm{c}}_\\textsc{map}$ of this <span style=\"font-variant:small-caps;\">gb-ppm</span> can be obtained leveraging the Bregman k-means algorithm [@Banerjee2005]. ", "This leads to very efficient computational routines for point estimation and enables uncertainty quantification for a wide class of existing methods, including the k-means example illustrated in Figure \\[fig:intro\\]. ", "First let us recall the definition of Bregman divergence.", "\n\nLet $\\varphi : \\mathds{X} \\rightarrow \\mathds{R}$ be a strictly convex function defined on a convex set $\\mathds{X} \\subseteq \\mathds{R}^d$, such that $\\varphi$ is differentiable on the relative interior of $\\mathds{X}$. A *Bregman divergence* is defined as $$\\mathcal{D}_\\varphi(\\bm{x}; \\bm{\\mu}) = \\varphi(\\bm{x}) - [\\varphi(\\bm{\\mu}) + (\\bm{x} - \\bm{\\mu})^\\intercal \\nabla\\varphi(\\bm{\\mu})],$$ for any $\\bm{x} \\in \\mathds{X}$ and any $\\bm{\\mu}$ in the relative interior of $\\mathds{X}$, where $\\nabla \\varphi(\\bm{\\mu})$ represents the gradient vector of $\\varphi$ evaluated at $\\bm{\\mu}$.\n\nA Bregman divergence $\\mathcal{D}_\\varphi(\\bm{x}; \\bm{\\mu})$ is always non-negative and it has a simple geometric interpretation: the “distance” between $\\bm{x}$ and $\\bm{\\mu}$ is measured as the difference between $\\varphi(\\bm{x})$ and the value of its tangent hyperplane at $\\bm{\\mu}$, evaluated at $\\bm{x}$. Many well known discrepancies are special cases of Bregman divergences, including the squared Euclidean distance, the Mahalanobis distance, and the <span style=\"font-variant:small-caps;\">kl</span>-divergence. ", "In the following, we employ Bregman divergences to construct a <span style=\"font-variant:small-caps;\">gb-ppm</span>.", "\n\n\\[Bregman\\] Let $\\pi_\\varphi(\\bm{c} \\mid \\lambda, \\bm{X})$ be a <span style=\"font-variant:small-caps;\">gb-ppm</span>. ", "We will say it has *Bregman cohesions* if $$\\pi_\\varphi(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\prod_{k=1}^K \\rho(C_k; \\lambda, \\bm{X}_k) = \\prod_{k=1}^K \\exp\\left\\{ - \\lambda \\sum_{i \\in C_k} \\: \\mathcal{D}_\\varphi(\\bm{x}_i; \\bar{\\bm{x}}_k)\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K,$$ where $\\mathcal{D}_\\varphi(\\bm{x}; \\bm{\\mu})$ is a Bregman divergence.", "\n\nNote that the arithmetic mean $\\bar{\\bm{x}}_k$ in the above definition is not an arbitrary choice, since it maximizes the associated cohesion function, that is, $\\bar{\\bm{x}}_k= \\arg \\max_{\\bm{\\mu}} \\exp\\{-\\lambda \\sum_{i \\in C_k} \\mathcal{D}_\\varphi(\\bm{x}_i; \\bm{\\mu})\\}$. Moreover, recall from Proposition \\[map\\] that $\\hat{\\bm{c}}_\\textsc{map}$ is obtained as the solution of the minimization problem $$\\min_{\\bm{c}\\: : \\: |\\bm{C}| = K}\\sum_{k=1}^K\\sum_{i \\in C_k} \\mathcal{D}_\\varphi(\\bm{x}_i; \\bar{\\bm{x}}_k).$$ The above optimization can be solved through the efficient Bregman k-means algorithm, which is recalled in Algorithm \\[Algorithm\\]. ", "Such a procedure can be regarded as a generalization of k-means and it entails essentially the same steps. ", "Moreover, it can be shown that Algorithm \\[Algorithm\\] monotonically decreases the loss function, and it reaches a local optimum in a finite number of steps [@Banerjee2005]. ", "Note also that in a <span style=\"font-variant:small-caps;\">gb-ppm</span> with Bregman cohesion the “generalized” centroids $\\bm{m}_1,\\dots,\\bm{m}_K$, defined in Section \\[sec:uq\\], correspond to the within-cluster means and therefore are straightforward to compute. ", "Indeed, for any $k=1,\\dots,K$ one has that $\\bm{m}_k = \\arg \\min_{\\bm{x} \\in \\mathds{X}}\\mathcal{D}_\\varphi(\\bm{x}; \\bar{\\bm{x}}_k) = \\bar{\\bm{x}}_k$, because $\\mathcal{D}_\\varphi(\\bm{x}; \\bar{\\bm{x}}_k) = 0$ if and only if $\\bm{x} = \\bar{\\bm{x}}_k$.\n\nFinally, we remark that the Bregman divergence $\\mathcal{D}_\\varphi(\\bm{x}_i; \\bar{\\bm{x}}_k)$ evaluated at $\\bar{\\bm{x}}_k$ is not always well-defined, because the arithmetic mean does not necessarily belong to the relative interior of $\\mathds{X}$. This typically occurs when the data have discrete support, e.g. when $x_{ij} \\in \\{0,1\\}$. This issue was not previously emphasized in @Banerjee2005, and it may lead to ill-behaved steps in Algorithm \\[Algorithm\\]. ", "We will address this difficulty in Section \\[sec:boundary\\] by relying on adjusted centroids $\\tilde{\\bm{x}}_k$.\n\nChoose $K$ and a set of initial centroids $\\bm{m}_1,\\dots,\\bm{m}_K$.\\\n$\\hat{\\bm{c}}_\\textsc{map} = (c_1, \\ldots, c_n)$.\\\n\nConnection with exponential dispersion families {#dispersion_families}\n-----------------------------------------------\n\nIt turns out that the <span style=\"font-variant:small-caps;\">gb-ppm</span> in Definition \\[Bregman\\] is deeply connected with mixtures of exponential dispersion models. ", "Such a class of distributions, introduced by @Jorgensen1987, is a generalization of regular exponential families and is indexed by a *dispersion parameter*. ", "We show that $\\lambda$ in Definition \\[Bregman\\] coincides with such a dispersion parameter. ", "This alternative probabilistic representation helps in the interpretation of the loss and in the elicitation of $\\lambda$, which is otherwise a difficult problem [@Bissiri2016; @Holmes2017]. ", "We refer to Section \\[sec:models\\] and Section \\[sec:illustration\\] for specific examples and default choices for $\\lambda$. We remark that such a connection holds only for *regular* Bregman divergences [@Banerjee2005], which are, however, the vast majority of the known cases.", "\n\n\\[dispersion\\] Let $\\pi(\\bm{x} \\mid \\lambda)$ be a density function on $\\mathds{X} \\subseteq \\mathds{R}^d$ indexed by $\\lambda > 0$ and let $\\Pi_\\lambda$ be its probability measure. ", "Then, the class of densities $$\\pi_{\\textsc{ed}}(\\bm{x} \\mid \\bm{\\theta},\\lambda) = \\pi(\\bm{x} \\mid \\lambda) e^{\\lambda [\\bm{\\theta}^\\intercal\\bm{x} - \\kappa(\\bm{\\theta})]}, \\qquad \\bm{\\theta} \\in \\Theta, \\quad \\lambda \\in \\Lambda,$$ is the *exponential dispersion family*, where $\\Theta \\times \\Lambda = \\{ (\\bm{\\theta},\\lambda) \\in \\mathds{R}^d \\times \\mathds{R}_+ \\setminus \\{0\\} : \\int_\\mathds{X} e^{\\lambda \\bm{\\theta}^\\intercal\\bm{x}}\\Pi_\\lambda({ \\,\\textup d}\\bm{x}) < \\infty\\}$.\n\nTo clarify the role of the parameters $\\lambda$ and $\\bm{\\theta}$, we recall the first two moments of an exponential dispersion random vector. ", "In particular, if $\\bm{x} \\sim \\pi_{\\textsc{ed}}(\\bm{x} \\mid \\bm{\\theta},\\lambda)$, then $$\\mathds{E}(\\bm{x}) = \\mu(\\bm{\\theta}), \\qquad \\text{Var}(\\bm{x}) = \\frac{1}{\\lambda} \\bm{V}(\\bm{\\theta}),$$ where the function $\\mu(\\cdot)$ is injective and $\\bm{V}(\\bm{\\theta})$ is a $d \\times d$ matrix not depending on $\\lambda$. Hence, there is a one-to-one correspondence between the so-called *natural parametrization* $\\bm{\\theta}$ and the *mean parametrization* $\\bm{\\mu} = \\mu(\\bm{\\theta})$, so that $\\bm{\\theta} = \\theta(\\bm{\\mu}) = \\mu^{-1}(\\bm{\\mu})$. Moreover, the parameters $\\lambda$ and $\\bm{\\theta}$ control the scale and the location of $\\bm{x}$, respectively. ", "The connection between exponential dispersion families and the <span style=\"font-variant:small-caps;\">gb-ppm</span> with Bregman cohesion is explained in the next Theorem.", "\n\n\\[profile\\_lik\\] Let $\\pi_\\textsc{ed}(\\bm{c} \\mid \\lambda, \\bm{X})$ be a <span style=\"font-variant:small-caps;\">gb-ppm</span> of the form $$\\begin{aligned}\n\\pi_\\textsc{ed}(\\bm{c} \\mid \\lambda, \\bm{X}) &\\propto \\prod_{k=1}^K\\prod_{i \\in C_k} \\pi_{\\textsc{ed}}(\\bm{x}_i \\mid \\hat{\\bm{\\theta}}_k, \\lambda) \\propto \\prod_{k=1}^K \\prod_{i \\in C_k} \\pi(\\bm{x}_i \\mid \\lambda) \\exp\\left\\{\\lambda [\\hat{\\bm{\\theta}}_k^\\intercal\\bm{x}_i - \\kappa(\\hat{\\bm{\\theta}}_k)]\\right\\}, \\quad \\bm{c} : |\\bm{C}| = K,\n\\end{aligned}$$ where each $\\hat{\\bm{\\theta}}_k = \\theta(\\bar{\\bm{x}}_k) =\\arg \\max_{\\bm{\\theta}} \\prod_{i \\in C_k} \\pi_{\\textsc{ed}}(\\bm{x}_i \\mid \\bm{\\theta}, \\lambda)$ is the maximum likelihood estimate of $\\bm{\\theta}_k$ for any given partition $\\bm{c}$. Then, there exists a <span style=\"font-variant:small-caps;\">gb-ppm</span> with Bregman cohesion as in Definition \\[Bregman\\] such that $$\\pi_\\textsc{ed}(\\bm{c} \\mid \\lambda, \\bm{X}) = \\pi_\\varphi(\\bm{c} \\mid \\lambda, \\bm{X}), \\qquad \\bm{c} : |\\bm{C}| = K,$$ for some suitable function $\\varphi$.\n\nBroadly speaking, any exponential dispersion family is associated to a <span style=\"font-variant:small-caps;\">gb-ppm</span> Bregman cohesion. ", "The Bregman divergences admitting such a representation are called *regular*. ", "The proof relies on the existence of a suitable function $\\varphi(\\bm{x})$ such that $$\\lambda \\bm{\\theta}^\\intercal\\bm{x} - \\lambda \\kappa(\\bm{\\theta}) = - \\lambda \\mathcal{D}_\\varphi(\\bm{x}; \\mu(\\bm{\\theta})) + \\lambda \\varphi(\\bm{x}),$$ a result which has been formally proved by @Banerjee2005. ", "Importantly, Theorem \\[profile\\_lik\\] also clarifies the role of $\\lambda$, which is proportional to the within-cluster precision. ", "In several practical cases $\\lambda$ may be set equal to a known constant, but it can be also estimated from the data; see Section \\[sec:models\\] for examples and practical remarks. ", "In addition, note that the maximum likelihood estimates $\\hat{\\bm{\\theta}}_k = \\theta(\\bar{\\bm{x}}_k)$ do not necessarily exist for all the partitions $\\bm{c}$. This indeed occurs whenever the associated Bregman divergence $\\mathcal{D}_\\varphi(\\bm{x}; \\bar{\\bm{x}}_k)$ is not well-defined. ", "These ill-behaved scenarios are discussed in Section \\[sec:boundary\\].", "\n\nFinally, a further consequence of Theorem \\[profile\\_lik\\] is that the <span style=\"font-variant:small-caps;\">gb-ppm</span> $\\pi_\\varphi(\\bm{c} \\mid \\lambda, \\bm{X})$ can be regarded as the Bayesian update of a *profile likelihood*, a likelihood function with nuisance parameters replaced by their maximum likelihood estimates. ", "Thus, $\\pi_\\textsc{ed}(\\bm{c} \\mid \\lambda, \\bm{X})$ may be seen also as an “approximate” Bayesian model. ", "Moreover, note that $\\hat{\\bm{c}}_\\textsc{map}$ can be actually interpreted as the maximum likelihood estimator. ", "Although this connection may be helpful for the choice of $\\lambda$, we stress again that we rely on the justification in Section \\[sec:bissiri\\], meaning that $\\pi_\\varphi(\\bm{c} \\mid \\lambda, \\bm{X})$ should be regarded as a proper update of beliefs rather than an approximate model.", "\n\nDiscrete data {#sec:boundary}\n-------------\n\nSome difficulties can arise in Bregman clustering procedures if the centroid $\\bar{\\bm{x}}_k$ is outside the relative interior of $\\mathds{X}$ or, equivalently, if the maximum likelihood estimate $\\hat{\\bm{\\theta}}_k$ does not exist. ", "Indeed, in such a case the Bregman divergence and the associated profile likelihood are not well-defined. ", "As previously mentioned, this issue is typically encountered whenever the data have discrete support.", "\n\nWhen the sample size is large enough, most of the partitions are well-defined and therefore one could disregard the cluster solutions that lead to ill-defined losses. ", "However, the representation of Theorem \\[profile\\_lik\\] suggests a more elegant fix. ", "Specifically, each $\\theta(\\bar{\\bm{x}}_k)$ can be replaced by a penalized maximum likelihood estimate $\\tilde{\\bm{\\theta}}_k$, so that $$\\tilde{\\bm{\\theta}}_k = \\arg \\max_{\\bm{\\theta}} \\pi(\\bm{\\theta} \\mid \\lambda) \\prod_{i \\in C_k} \\pi_{\\textsc{ed}}(\\bm{x}_i \\mid \\bm{\\theta}, \\lambda), \\qquad k=1,\\dots, K,$$ for a suitable penalty function $\\pi(\\bm{\\theta} \\mid \\lambda)$ that leads to well-defined estimates. ", "Equivalently, we can replace the arithmetic means $\\bar{\\bm{x}}_k$ with the adjusted centroids $\\tilde{\\bm{x}}_k = \\mu(\\tilde{\\bm{\\theta}}_k)$, so that $\\tilde{\\bm{\\theta}}_k = \\theta(\\tilde{\\bm{x}}_k)$. To preserve the <span style=\"font-variant:small-caps;\">gb-ppm</span> structure, we require that the estimates $\\tilde{\\bm{\\theta}}_k$ do not depend on $\\lambda$.\n\nIn exponential dispersion families, a convenient choice for the penalty function $\\pi(\\bm{\\theta} \\mid \\lambda)$ is given by @Firth1993. ", "In such a case, the estimates $\\theta(\\tilde{\\bm{x}}_k)$ have improved asymptotic properties and are typically easy to compute. ", "Moreover, the penalty function $\\pi(\\bm{\\theta} \\mid \\lambda)$ can be interpreted as the Jeffrey’s prior associated to $\\pi_{\\textsc{ed}}(\\bm{x} \\mid \\bm{\\theta}, \\lambda)$. Since the adjusted centroids $\\tilde{\\bm{x}}_k$ do not depend on $\\lambda$, we then obtain the following <span style=\"font-variant:small-caps;\">gb-ppm</span> $$\\label{adjusted_gbppm}\n\\pi_\\varphi(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\prod_{k=1}^K \\prod_{i \\in C_k} \\pi_{\\textsc{ed}}(\\bm{x}_i \\mid \\theta(\\tilde{\\bm{x}}_k), \\lambda) \\propto \\prod_{k=1}^K \\exp\\left\\{ - \\lambda \\sum_{i \\in C_k} \\: \\mathcal{D}_\\varphi(\\bm{x}_i; \\tilde{\\bm{x}}_k)\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K,$$ which is well-defined for any choice of the partition $\\bm{c}$. Importantly, the $\\hat{\\bm{c}}_\\textsc{map}$ solution can be still obtained using Algorithm \\[Algorithm3\\], which is a simple modification of Bregman k-means that preserves its main properties. ", "This is clarified in the following Proposition.", "\n\nChoose $K$ and a set of initial centroids $\\bm{m}_1,\\dots,\\bm{m}_K$.\\\n$\\hat{\\bm{c}}_\\textsc{map} = (c_1, \\ldots, c_n)$.\\\n\n\\[monotone\\] The Bregman k-means with adjusted centroids in Algorithm \\[Algorithm3\\] monotonically decreases the loss function $\\ell(\\bm{c};\\bm{X}) = \\sum_{k=1}^K\\sum_{i \\in C_k} \\mathcal{D}_\\varphi(\\bm{x}_i; \\tilde{\\bm{x}}_k)$ for any $\\bm{c} : |\\bm{C}| =K$. Moreover, Algorithm \\[Algorithm3\\] terminates in a finite number of steps at a partition that is locally optimal.", "\n\nCohesions based on pairwise dissimilarities {#sec:distance}\n===========================================\n\nWe now specialize the general <span style=\"font-variant:small-caps;\">gb-ppm</span> formulation in a different direction, and we describe a family of cohesion functions based on pairwise dissimilarities. ", "Such a family may offer a more robust clustering framework compared to Bregman cohesions. ", "The robustness properties of this <span style=\"font-variant:small-caps;\">gb-ppm</span> will depend on the chosen dissimilarity measure. ", "Within the algorithmic framework, the closest relative to this <span style=\"font-variant:small-caps;\">gb-ppm</span> approach is the so-called k-medoids or <span style=\"font-variant:small-caps;\">pam</span> algorithm [@Kaufman1990; @Hastie2008], which indeed may be used to mitigate the drawbacks of k-means. ", "As we shall see, this class is also closely related to the approximate Bayesian model of @Duan2019.", "\n\nLet us assume that the covariate space is $\\mathds{X} = \\mathds{R}^d$ and let $||\\bm{x}||_p = (|x_1|^p + \\cdots + |x_d|^p)^{1/p}$ be the $L^p$ norm, for any $p \\ge 1$ and $\\bm{x} \\in \\mathds{R}^d$. Then, a general measure of dissimilarity is $$\\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p), \\qquad \\bm{x}_i, \\bm{x}_{i'} \\in \\mathds{R}^d,$$ for some increasing function $\\gamma : \\mathds{R}_+ \\rightarrow \\mathds{R}_+$ such that $\\gamma(0) = 0$. In general $\\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p)$ is not a metric on $\\mathds{R}^d$, although this could be the case for specific choices of $\\gamma$. For example, with $\\gamma(x) = x^{1/p}$ one obtains the Minkowski distance. ", "The squared Euclidean distance is recovered when $\\gamma(x) = x$ and $p = 2$. Recalling the definition of a <span style=\"font-variant:small-caps;\">gb-ppm</span>, we then seek a discrepancy function measuring the distance of the $i$th unit from the $k$th cluster. ", "We consider the so-called average dissimilarity $$\\label{average}\n\\mathcal{D}_\\gamma(\\bm{x}_i; \\bm{X}_k) = \\frac{1}{n_k} \\sum_{i' \\in C_k} \\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p),$$ which leads to the following novel <span style=\"font-variant:small-caps;\">gb-ppm</span>.", "\n\n\\[pairwise\\] Let $\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X})$ be a <span style=\"font-variant:small-caps;\">gb-ppm</span> with covariate space $\\mathds{X} = \\mathds{R}^d$. We will say it has *average dissimilarity cohesions* if $$\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\prod_{k=1}^K \\rho(C_k; \\lambda, \\bm{X}_k) = \\prod_{k=1}^K \\exp\\left\\{ - \\frac{\\lambda}{2}\\sum_{i \\in C_k} \\frac{1}{n_k}\\sum_{i' \\in C_k} \\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p)\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K,$$ with $p \\ge 1$ and $\\gamma : \\mathds{R}_+ \\rightarrow \\mathds{R}_+$ an increasing function such that $\\gamma(0) = 0$.\n\nRandomly allocate the indicators $c_1,\\dots,c_n$ into $K$ sets.\\\n$\\hat{\\bm{c}}_\\textsc{map} = (c_1, \\ldots, c_n)$.\\\n\nPosterior inference in a <span style=\"font-variant:small-caps;\">gb-ppm</span> with average dissimilarity cohesions is not as fast as in the Bregman divergence case, but it is still quite efficient. ", "Indeed, one needs to evaluate all the distinct $n(n-1)/2$ pairwise dissimilarities, which could be a computational bottleneck. ", "However, if these values are pre-computed and stored, then inference can be easily conducted. ", "Gibbs sampling is performed by iteratively drawing values from the full-conditionals in Theorem \\[GibbsSampling\\]. ", "Moreover, an efficient routine for finding $\\hat{\\bm{c}}_\\textsc{map}$ is available and is described in Algorithm \\[Algorithm4\\]. ", "We call this novel method *k-dissimilarities*, which might be of independent interest. ", "Such an algorithm retains the main properties of the k-means algorithm, as clarified in the following Proposition.", "\n\n\\[monotone2\\] The iterative allocations in the k-dissimilarities Algorithm \\[Algorithm4\\] monotonically decrease the loss function $\\ell(\\bm{c};\\bm{X}) = \\sum_{k=1}^K\\sum_{i \\in C_k} \\mathcal{D}_\\gamma(\\bm{x}_i; \\bm{X}_k)$ for any $\\bm{c} : |\\bm{C}| =K$. Moreover, Algorithm \\[Algorithm4\\] terminates in a finite number of steps at a partition that is locally optimal.", "\n\nThe results of Proposition \\[monotone2\\] are not surprising, because at each step the k-dissimilarities algorithm locally minimizes the loss function. ", "Moreover, note that the discrepancy $\\sum_{i \\in C_k} \\mathcal{D}_\\gamma(\\bm{x}_i; \\bm{X}_k)$ should not be re-computed at each step, thanks to a recursive formula. ", "Such a recursion, described in Appendix \\[app1\\], is extremely useful also in the Gibbs sampling implied by Theorem \\[GibbsSampling\\].", "\n\nConnection with spherical distributions {#composite_likelihoods}\n---------------------------------------\n\nIt turns out that <span style=\"font-variant:small-caps;\">gb-ppm</span>s with pairwise dissimilarities are deeply connected with $L^p$ spherical distributions [@Gupta1997]. ", "Such a class of random vectors in $\\mathds{R}^d$ is a generalization of spherical distributions in $L^2$; see @Fang1990 for an overview. ", "Paralleling the discussion of Section \\[dispersion\\_families\\], these probabilistic connections improve the interpretability of the model and facilitate the choice or the estimation of $\\lambda$. This will be discussed in detail in Section \\[sec:models\\]. ", "We begin our discussion by providing the definition of $L^p$ spherical distributions.", "\n\nA random vector $\\bm{x} \\in \\mathds{R}^d$ follows a $L^p$ spherical distribution if its density function can be written as $\\pi_\\textsc{sp}(\\bm{x}) = g(||\\bm{x}||_p^p)$ for some measurable function $g : \\mathds{R}_+ \\rightarrow \\mathds{R}_+$.\n\nThe class of $L^p$ spherical distributions is very general as it includes the multivariate Gaussian, the multivariate Laplace, and the multivariate Student’s t, among others. ", "Such a family is indexed by the function $g$, which is sometimes called the *density generator*. ", "The connection with the <span style=\"font-variant:small-caps;\">gb-ppm</span> of Definition \\[pairwise\\] is clarified in the following Theorem.", "\n\n\\[pairwise\\_lik\\] Let $\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X})$ be a <span style=\"font-variant:small-caps;\">gb-ppm</span> with average dissimilarities as in Definition \\[pairwise\\]. ", "If $$\\int_{\\mathds{R}_+} r^{d-1} \\exp\\left\\{-\\frac{\\lambda}{2} \\gamma(r^p)\\right\\}{ \\,\\textup d}r < \\infty,$$ then there exists an $L^p$ spherical distribution on $\\mathds{R}^d$ such that $$\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\prod_{k=1}^K \\prod_{i \\in C_k} \\left[\\prod_{i' \\in C_k} \\pi_\\textsc{sp}(\\bm{x}_i - \\bm{x}_{i'} \\mid \\lambda)\\right]^{1/n_k},$$ where $\\pi_\\textsc{sp}(\\bm{x}_i - \\bm{x}_{i'} \\mid \\lambda) \\propto \\exp\\left\\{-\\lambda/2\\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p)\\right\\}$ for any $i\\in C_k$ and $i' \\in C_k$.\n\nHence, if a simple integrability condition on the $\\gamma$ function holds true, then the <span style=\"font-variant:small-caps;\">gb-ppm</span> of Definition \\[pairwise\\] has a simple probabilistic interpretation. ", "In addition, Theorem \\[pairwise\\_lik\\] implies that $\\exp\\{-\\lambda\\gamma(x)\\}$ is proportional to a density generator for $L^p$ spherical distributions. ", "The role of $\\lambda$ will depend on the specific choices of $\\gamma$, but we will show in Section \\[sec:models\\] that $\\lambda$ can often be interpreted as a scale parameter.", "\n\nImportantly, Theorem \\[pairwise\\_lik\\] highlights that a <span style=\"font-variant:small-caps;\">gb-ppm</span> with average dissimilarities can be interpreted as the Bayesian update of a *pairwise difference likelihood* [@Lele2002; @Varin2011], a special instance of composite likelihood. ", "Suppose that, conditionally on the cluster indicators, the observations follow some location family of distributions, namely $$(\\bm{x}_i \\mid \\bm{\\mu}_k, \\lambda, c_i = k) \\overset{\\textup{iid}}{\\sim} \\pi\\left(\\bm{x} - \\bm{\\mu}_k \\mid \\lambda \\right), \\qquad i \\in C_k, \\quad k=1,\\dots,K,$$ where $\\bm{\\mu}_k \\in \\mathds{R}^d$. Following @Lele2002 [@Varin2011] one could consider a within-cluster pairwise difference likelihood, which is based on the differences $\\bm{x}_i - \\bm{x}_{i'}$, for any $i \\in C_k$ and $i' \\in C_k$. Note that the within-cluster differences $\\bm{x}_i - \\bm{x}_{i'}$ are identically distributed and they do not depend on the location parameter $\\bm{\\mu}_k$. $L^p$ spherical distributions are indeed an appealing and very general modeling choice for $\\bm{x}_i - \\bm{x}_{i'}$, e.g. because they are symmetric around $0$, a natural requirement in this setting. ", "Hence, the associated pairwise difference likelihood is proportional to $$\\pi_\\textsc{diff}(\\bm{X} \\mid \\bm{c}, \\lambda) \\propto \\prod_{k=1}^K \\prod_{i \\in C_k} \\left[\\prod_{i' \\in C_k}\\pi_\\textsc{sp}(\\bm{x}_i - \\bm{x}_{i'} \\mid \\lambda)\\right]^{1/n_k},$$ where the exponent $1/n_k$ is a correction that deflates the likelihood so that it has the usual asymptotic order. ", "The key advantage of this kind of composite likelihood is that the location parameters have been removed. ", "As may already be clear, the <span style=\"font-variant:small-caps;\">gb-ppm</span> in Definition \\[pairwise\\] corresponds to the Bayesian update of $\\pi_\\textsc{diff}(\\bm{X} \\mid \\bm{c}, \\lambda)$ if a suitable integrability condition holds. ", "As discussed before, note that we are not regarding $\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X})$ as an approximate Bayesian model based on composite likelihoods as in @Pauli2011 or @Aliverti2020. ", "Instead, we stress once again that we rely on the coherent update of beliefs presented in Section \\[sec:bissiri\\].", "\n\nModeling examples {#sec:models}\n=================\n\nIn this section we present specific <span style=\"font-variant:small-caps;\">gb-ppm</span>s belonging to the classes discussed in Sections \\[sec:Bregman\\] and \\[sec:distance\\]. ", "In doing so, we also address the issue of either specifying or estimating the value of $\\lambda$. We suggest default strategies based on the theoretical findings of the previous sections. ", "We emphasize that the models presented here were chosen for the sake of illustration. ", "Indeed, we think that a core contribution of our approach is the generality of <span style=\"font-variant:small-caps;\">gb-ppm</span>s, and therefore we are not claiming the superiority of these specific cases over alternative generalized Bayes models. ", "Instead, we aim at providing some guidelines for their construction, usage and interpretation.", "\n\nSquared euclidean distance {#sec:squared_euclidean}\n--------------------------\n\nThe first <span style=\"font-variant:small-caps;\">gb-ppm</span> we describe relies on the quadratic loss given in equation , which is closely related to k-means. ", "Given the central role the k-means algorithm plays in clustering problems, we believe the associated <span style=\"font-variant:small-caps;\">gb-ppm</span> deserves special attention. ", "If $\\mathds{X} = \\mathds{R}^d$, then it is easy to show that the loss  is a Bregman divergence, implying that the associated <span style=\"font-variant:small-caps;\">gb-ppm</span> is $$\\label{kmeans_gbppm}\n\\pi_\\varphi(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\prod_{k=1}^K \\exp\\left\\{- \\lambda \\sum_{i \\in C_k} ||\\bm{x}_i - \\bar{\\bm{x}}_k||_2^2\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K,$$ with $\\varphi(\\bm{x}) = ||\\bm{x}||_2^2$. The above model can also be regarded as a <span style=\"font-variant:small-caps;\">gb-ppm</span> with pairwise dissimilarities. ", "Recall the well known identity $\\lambda \\sum_{i \\in C_k} ||\\bm{x}_i - \\bar{\\bm{x}}_k||_2^2 = \\lambda/2 \\sum_{i \\in C_k} n_k^{-1} \\sum_{i' \\in C_k} ||\\bm{x}_i - \\bm{x}_{i'}||_2^2$. Then, with $\\gamma(x) = x$ and $p=2$ we can equivalently let $$\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\prod_{k=1}^K \\exp\\left\\{ - \\frac{\\lambda}{2}\\sum_{i \\in C_k} \\frac{1}{n_k}\\sum_{i' \\in C_k} ||\\bm{x}_i - \\bm{x}_{i'}||_2^2\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K.$$ Hence, the point estimate $\\hat{\\bm{c}}_\\textsc{map}$ can be either obtained through Algorithm \\[Algorithm\\] or \\[Algorithm4\\]. ", "The former coincides with the k-means algorithm of @Lloyd1982, whereas the latter is essentially a variation of the algorithm of @Hartigan1979.", "\n\nThe k-means <span style=\"font-variant:small-caps;\">gb-ppm</span> therefore inherits the probabilistic interpretation of Theorems \\[profile\\_lik\\] and \\[pairwise\\_lik\\]. ", "Specifically, the associated profile and composite likelihoods are consistent with the generative mechanism $$\\label{kmeans_profile}\n(\\bm{x}_i \\mid \\bm{\\mu}_k, \\lambda, c_i = k) \\overset{\\textup{iid}}{\\sim} \\mbox{N}(\\bm{\\mu}_k, (2\\lambda)^{-1} I_d), \\qquad i \\in C_k, \\quad k=1,\\dots,K,$$ where $\\mbox{N}(\\bm{\\mu},\\bm{\\Sigma})$ denotes a $d$-dimensional multivariate Gaussian with mean $\\bm{\\mu}$ and covariance matrix $\\bm{\\Sigma}$. The multivariate Gaussian distribution in  is indeed an exponential dispersion family. ", "In addition, it follows that the $L^2$ spherical distribution characterizing the pairwise distances $\\bm{x}_i - \\bm{x}_{i'}$ must be a multivariate Gaussian, namely $$\\label{kmeans_diff}\n(\\bm{x}_i - \\bm{x}_{i'} \\mid \\lambda, c_i = k, c_{i'} = k) \\sim \\mbox{N}\\left(\\bm{0}, \\lambda^{-1} I_d\\right),$$ for any $i \\in C_k$, $i' \\in C_k$ and $k=1,\\dots,K$. Note that the profile likelihood obtained from equation  coincides with the pairwise difference likelihood based on equation .", "\n\nExploiting these probabilistic connections, we discuss a reasonable default strategy for the estimation of $\\lambda$. Following @Bissiri2016, we incorporate $\\lambda$ in the loss function and we specify a hierarchical prior $\\pi(\\lambda)$. The balance between the joint prior $\\pi(\\bm{c})\\pi(\\lambda)$ and the joint loss $\\ell(\\bm{c},\\lambda;\\bm{X})$ is regulated by a parameter $\\tilde{\\lambda} > 0$, whose choice is simpler than the one for $\\lambda$. More precisely, we consider the following joint Gibbs posterior $$\\label{gibbs_joint}\n\\begin{aligned}\n\\pi(\\bm{c}, \\lambda \\mid \\tilde{\\lambda}, \\bm{X}) &\\propto \\pi(\\bm{c}) \\pi(\\lambda) \\exp\\{- \\tilde{\\lambda} \\ell(\\bm{c},\\lambda;\\bm{X}) \\}, \\quad \\text{with} \\quad \\ell(\\bm{c},\\lambda;\\bm{X}) = \\lambda\\ell(\\bm{c};\\bm{X}) - \\xi \\log(\\lambda), \\\\\n\\end{aligned}$$ with $\\bm{c} : |\\bm{C}| = K$ and for some $\\xi \\ge 0$. Theorems \\[profile\\_lik\\] and \\[pairwise\\_lik\\] suggest that joint inference about $\\bm{c}$ and $\\lambda$ should be based on the prior updated with the appropriate pseudo-likelihood as $$\\label{kmeans_joint}\n\\pi(\\bm{c}, \\lambda \\mid \\bm{X}) \\propto \\pi(\\lambda) \\lambda^{nd/2} \\prod_{k=1}^K \\exp\\left\\{- \\lambda \\sum_{i \\in C_k} ||\\bm{x}_i - \\bar{\\bm{x}}_k||_2^2\\right\\},\\qquad \\bm{c} : |\\bm{C}| = K.$$ Comparing equation  with we see that $\\xi = nd/2$ and $\\tilde{\\lambda} = 1$ are natural default choices. ", "Note that the additional term $\\lambda^{nd/2}$ follows from Theorems \\[profile\\_lik\\] and \\[pairwise\\_lik\\]. ", "Moreover, if we let $\\lambda \\sim \\textsc{gamma}(a_\\lambda, b_\\lambda)$ a priori, then the full conditional $\\pi(\\lambda \\mid \\bm{c}, \\bm{X})$ is still a Gamma with updated parameters $a_\\lambda + nd/2$ and $b_\\lambda + \\sum_{k=1}^K \\sum_{i \\in C_k} ||\\bm{x}_i - \\bar{\\bm{x}}_k||_2^2$, leading to a simple Gibbs sampling step.", "\n\nWe remark that the choices $\\xi = nd/2$ and $\\tilde{\\lambda} = 1$ represent a reasonable default, but alternatives might be considered. ", "For example, in the same spirit of @Miller2019, one could let $\\xi = nd/2$ and $0 < \\tilde{\\lambda} < 1$, thus obtaining a variant of the joint model  which deflates the importance of the pseudo-likelihood. ", "The parameter $\\tilde{\\lambda}$ can be then selected either subjectively or following the guidelines presented in their paper.", "\n\nMinkowski distance {#sec:Minkowski}\n------------------\n\nAs a more robust alternative to the k-means case, we present a <span style=\"font-variant:small-caps;\">gb-ppm</span> based on pairwise dissimilarities, which essentially replaces the squared Euclidean metric with the $L^p$ distance. ", "For $p=1$ this makes the clustering results less sensitive to outliers. ", "More precisely, if $\\gamma(x) = x^{1/p}$ then $\\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p) = ||\\bm{x}_i - \\bm{x}_{i'}||_p$ is the Minkowski distance, which is also called Manhattan distance when $p=1$. Then, $$\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X}) \\propto \\prod_{k=1}^K \\exp\\left\\{ - \\frac{\\lambda}{2}\\sum_{i \\in C_k} \\frac{1}{n_k}\\sum_{i' \\in C_k} ||\\bm{x}_i - \\bm{x}_{i'}||_p\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K.$$ This specific <span style=\"font-variant:small-caps;\">gb-ppm</span> is closely related to a special case of @Duan2019, the only differences being the prior $\\pi(\\bm{c})$ and the condition $\\bm{c} : |\\bm{C}| = K$. However, we remark that we regard $\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X})$ as a generalized Bayes posterior rather than a “partially specified” Bayesian model. ", "Within our framework, the derivation of $\\pi_\\gamma(\\bm{c} \\mid \\lambda, \\bm{X})$ is very natural, being the combination of the factorized loss  and the arithmetic average given in . ", "Hence, when $\\gamma(x) = x^{1/p}$ our contribution leads to an alternative derivation of the model of @Duan2019, to the development of novel algorithms, and to an important probabilistic interpretation based on composite likelihoods.", "\n\nIt is easy to check that the condition $\\int_{\\mathds{R}_+} r^{d-1} e^{-\\lambda/2 r}{ \\,\\textup d}r < \\infty$ required by Theorem \\[pairwise\\_lik\\] is satisfied. ", "Moreover, the $L^p$ spherical distribution associated to each pairwise difference $\\bm{x}_i - \\bm{x}_{i'}$, for any $i \\in C_k, i' \\in C_k$, has density $$\\pi_\\textsc{sp}(\\bm{x}_i - \\bm{x}_{i'} \\mid \\lambda) = \\frac{p^{d-1} }{2^d \\Gamma(1/p)^d} \\frac{\\Gamma(d/p)}{\\Gamma(d)}\\left(\\frac{\\lambda}{2}\\right)^d \\exp\\left\\{ - \\frac{\\lambda}{2}||\\bm{x}_i - \\bm{x}_{i'}||_p \\right\\}.$$ Thus, $\\lambda$ can be interpreted as a scale parameter. ", "The normalizing constant of the above density is related to the integral $\\int_{\\mathds{R}_+} r^{d-1} e^{-\\lambda/2 r}{ \\,\\textup d}r$ [@Gupta1997], which has an explicit solution. ", "Moreover, for any $p > 1$, the components of the above random vector are dependent, but when $p = 1$ we recover the law of $d$ independent Laplace distributions. ", "Finally, simple properties of $L^p$ spherical distributions imply that $||\\bm{x}_i - \\bm{x}_{i'}||_p \\sim \\textsc{gamma}(d, \\lambda/2)$ for any $p \\ge 1$.\n\nWe again suggest to estimate $\\lambda$ from the data. ", "Specifically, joint inference on $\\lambda$ and $\\bm{c}$ can be conducted as in equation , i.e. through a hierarchical approach. ", "On the other hand, Theorem \\[pairwise\\_lik\\] suggests that joint inference should be based on the update of $$\\pi(\\bm{c}, \\lambda \\mid \\bm{X}) \\propto \\pi(\\lambda) \\lambda^{n d} \\prod_{k=1}^K \\exp\\left\\{ - \\frac{\\lambda}{2}\\sum_{i \\in C_k} \\frac{1}{n_k}\\sum_{i' \\in C_k} ||\\bm{x}_i - \\bm{x}_{i'}||_p\\right\\}, \\qquad \\bm{c} : |\\bm{C}| = K,$$ where the additional term $\\lambda^{nd}$ follows from the pairwise difference likelihood representation. ", "This leads to the default values $\\xi = n d$ and $\\tilde{\\lambda} = 1$. The prior $\\lambda \\sim \\textsc{gamma}(a_\\lambda, b_\\lambda)$ is a computationally convenient choice also in this case, since it leads to a conjugate Gibbs sampling step.", "\n\nKullback-Leibler divergence {#sec:kl_model}\n---------------------------\n\nThe last <span style=\"font-variant:small-caps;\">gb-ppm</span> model we describe involves discrete data. ", "Specifically, suppose $\\mathds{X} = \\{0,1\\}^d$, meaning that $\\bm{x}_i =(x_{i1},\\dots,x_{id})^\\intercal$ is a collection of binary indicators. ", "The nature of the data leads to ill-behaved losses, as discussed in Section \\[sec:boundary\\]. ", "To overcome these difficulties, we rely on a <span style=\"font-variant:small-caps;\">gb-ppm</span> with adjusted centroids $\\tilde{\\bm{x}}_k = (\\tilde{x}_{k1},\\dots,\\tilde{x}_{kd})^\\intercal$, as in equation , and we let $$\\label{kl_gbppm}\n\\begin{aligned}\n\\pi_\\varphi(\\bm{c} \\mid \\lambda=1, \\bm{X}) &\\propto \\prod_{k=1}^K \\exp\\left\\{- \\sum_{i \\in C_k}\\sum_{j=1}^d \\textsc{kl}(p_{ij} \\mid\\mid \\tilde{p}_{kj})\\right\\}, \\quad \\bm{c} :|\\bm{C}| = K,\\\\\n &\\propto \\prod_{k=1}^K \\prod_{i \\in C_k} \\exp\\left\\{\\sum_{j=1}^d\\left[ x_{ij} \\log\\left(\\frac{\\tilde{x}_{kj}}{1 - \\tilde{x}_{kj}}\\right) + \\log(1 - \\tilde{x}_{kj})\\right]\\right\\}, \\quad \\bm{c} :|\\bm{C}| = K, \n\\end{aligned}$$ where each $p_{ij} = (x_{ij},1- x_{ij})^\\intercal$ and $\\tilde{p}_{kj} = (\\tilde{x}_{kj},1-\\tilde{x}_{kj})^\\intercal$ are vectors of probabilities. ", "The Kullback-Leibler divergence is well-defined even if $p_{ij}$ are degenerate probabilities. ", "The dispersion parameter $\\lambda = 1$ is fixed. ", "Considering equation , the kernel $\\pi_{\\textsc{ed}}(\\bm{x} \\mid \\bm{\\theta},\\lambda=1)$ is that of $d$ independent Bernoulli random variables, linking this <span style=\"font-variant:small-caps;\">gb-ppm</span> with latent class models [@Lazarsfeld1968]. ", "The adjusted centroids are obtained exploiting the @Firth1993 correction for $\\theta(\\tilde{\\bm{x}}_k)$, which leads to $$\\tilde{x}_{kj} = \\frac{n_k}{n_k +1}\\bar{x}_{kj} + \\frac{1}{2}\\frac{1}{n_k +1}, \\qquad j=1,\\dots,d, \\quad k=1,\\dots,K,$$ where $\\bar{x}_{kj}$ denotes the within-cluster proportion. ", "Boundary issues are avoided because $\\tilde{x}_{kj} \\in (0,1)$. Equation  relies on the pseudo-likelihood representation from Theorem \\[profile\\_lik\\], which implies that $\\lambda = 1$.\n\nIllustrations {#sec:illustration}\n=============\n\nIn this section, we further illustrate <span style=\"font-variant:small-caps;\">gb-ppm</span>s through synthetic and real data examples. ", "For simplicity, we focus on the models described in Section \\[sec:models\\]. ", "In Section \\[sec:sim1\\] we illustrate the uncertainty quantification tools described in Section \\[sec:uq\\] on a synthetic dataset and compare the results with “oracle” clustering probabilities associated with the true generative mechanism. ", "In Section \\[sec:sim2\\] we show that <span style=\"font-variant:small-caps;\">gb-ppm</span>s with pairwise dissimilarities improve on robustness compared with k-means based approaches. ", "In Section \\[sec:realdata\\] we illustrate the performance of a <span style=\"font-variant:small-caps;\">gb-ppm</span> model for binary data with a medical application considered in @Agresti2002 [Chap. ", "13]. ", "Moreover, we show that alternative point estimates, such as those presented in @Wade2018, can be more appropriate than the <span style=\"font-variant:small-caps;\">map</span>.", "\n\nUncertainty quantification and oracle clustering {#sec:sim1}\n------------------------------------------------\n\nIn this first simulation study we focus on the <span style=\"font-variant:small-caps;\">gb-ppm</span> with squared Euclidean loss described in Section \\[sec:squared\\_euclidean\\]. ", "We consider $n = 200$ observations evenly divided in $K = 4$ clusters, each having $n_1 = \\cdots = n_4 = 50$ data points. ", "Within each partition we let $$\\label{sim1}\n(\\bm{x}_i \\mid \\bm{\\mu}_k, \\sigma^2, c_i = k) \\overset{\\textup{iid}}{\\sim} \\mbox{N}\\left(\\bm{\\mu}_k, \\sigma^2 I_2\\right), \\qquad i \\in C_k, \\quad k=1,\\dots,K,$$ with $\\bm{\\mu}_1 = (-2,-2)$, $\\bm{\\mu}_2 = (-2,2)$, $\\bm{\\mu}_3 = (2,-2)$, and $\\bm{\\mu}_4 = (2,2)$. The within-cluster variance $\\sigma^2$ takes the values reported in Table \\[tab1\\]. ", "We let $K = 4$ for simplicity and we use the default strategy described in Section \\[sec:squared\\_euclidean\\], with $\\xi = nd/2$, $\\tilde{\\lambda}=1$, and $\\pi(\\lambda) \\propto 1$. Points estimates are obtained using k-means, and for uncertainty quantification we run the Gibbs sampler of Theorem \\[GibbsSampling\\] starting at the k-means estimate. ", "We obtained $5,000$ posterior samples, after a burn-in of $1,000$ iterations. ", "The traceplot of the loss function displays excellent mixing and no evidence against convergence. ", "Horizontal bounds were computed through the `mcclust.ext` R package.", "\n\n![", "Synthetic data points generated from equation  with $\\sigma^2 = 1.5$. Left plot: colors and shapes represent different clusters in the k-means point estimate $\\hat{\\bm{c}}_\\textsc{map}$. Right plot: colors and shapes represent different clusters in the horizontal bound $\\hat{\\bm{c}}_{1-\\alpha,\\textsc{map}}$, with $\\alpha = 0.05$. \\[fig:sim1\\_hb\\]](Sim1_hb){width=\"\\textwidth\"}\n\nWe display in Figure \\[fig:sim1\\_hb\\] the generated data points when $\\sigma^2 = 1.5$, together with the k-means estimate $\\hat{\\bm{c}}_\\textsc{map}$ and its horizontal bound $\\hat{\\bm{c}}_{1-\\alpha,\\textsc{map}}$, with $\\alpha = 0.05$. Recall that $\\hat{\\bm{c}}_{1-\\alpha,\\textsc{map}}$ provides the clustering that is furthest from the k-means solution $\\hat{\\bm{c}}_\\textsc{map}$ while still falling within a 95% credible region around $\\hat{\\bm{c}}_\\textsc{map}$. Figure \\[fig:sim1\\_hb\\] shows that $\\hat{\\bm{c}}_{1-\\alpha,\\textsc{map}}$ is roughly similar to $\\hat{\\bm{c}}_\\textsc{map}$ but many of the points near the cluster boundaries are assigned to different groups. ", "As evidenced in Table \\[tab1\\], the variation of information between the <span style=\"font-variant:small-caps;\">map</span> and its horizontal bound depends on the variability of the data, as expected. ", "In particular, more dispersed observations are associated with wider credible intervals, and vice versa.", "\n\n![", "Co-clustering probabilities $\\bm{S}_\\textsc{oracle}$ and $\\bm{S}$ for the oracle model and the squared Euclidean <span style=\"font-variant:small-caps;\">gb-ppm</span>. ", "Data points are generated from equation , with $\\sigma^2 = 1.5$ and ordered according to the true partition $\\bm{c}_0$.\\[fig:sim1\\_co\\]](Sim1_Co){width=\"\\textwidth\"}\n\n![", "Synthetic data points generated from equation  with $\\sigma^2 = 1.5$. Colors represent the misclassification probabilities obtained from the oracle co-clustering matrix $\\bm{S}_\\textsc{oracle}$ and squared Euclidean co-clustering matrix $\\bm{S}$, respectively. ", "\\[fig:sim1\\_miscl\\]](Sim1_miscl){width=\"\\textwidth\"}\n\nFrom equation  one can obtain *misclassification probabilities* $1 - s_{i \\hat{i}^*}$ of a point estimate $\\hat{\\bm{c}}$, where $\\hat{i}^* = i^*_{\\hat{c}_i}$ denotes the medoid of the cluster to which the $i$th unit has been allocated. ", "The quantity $1 - s_{i \\hat{i}^*}$ approximates the probability that the $i$th unit is allocated to a cluster different from $\\hat{c}_i$. To assess the appropriateness of these indicators, we compare them with a gold standard, which we call oracle clustering. ", "Specifically, the oracle distribution associated to the generative mechanism  is $$\\pi_\\textsc{oracle}(\\bm{c} \\mid \\bm{\\mu}_1,\\dots,\\bm{\\mu}_K, \\sigma^2, \\bm{X}) \\propto \\prod_{i=1}^n \\prod_{k=1}^K \\mbox{N}(\\bm{x}_i \\mid \\bm{\\mu}_k, \\sigma^2I_2)^{\\mathds{1}(c_i = k)},$$ where $\\mathds{1}(\\cdot)$ denotes the indicator function, $\\bm{\\mu}_k$ and $\\sigma^2$ are defined as in , and $\\mbox{N}(\\bm{x} \\mid \\bm{\\mu}, \\bm{\\Sigma})$ is the multivariate Gaussian density function. ", "The oracle clustering distribution represents the posterior distribution if the data generating process were known. ", "From $\\pi_\\textsc{oracle}(\\bm{c} \\mid \\bm{X})$ we derive the corresponding point estimate $\\hat{\\bm{c}}_{\\textsc{oracle}} = \\arg \\max_{\\bm{c}} \\pi_\\textsc{oracle}(\\bm{c} \\mid \\bm{X})$, the co-clustering matrix $\\bm{S}_\\textsc{oracle}$ having entries $s_{ii',\\textsc{oracle}}$ for $i,i'=1,\\dots,n$, and the associated misclassification probabilities.", "\n\n $\\sigma^2$ $\\{n(n-1)/2\\}^{-1}\\sum_{i < i'}|s_{ii'} - s_{ii',\\textsc{oracle}}|$ $d_\\textsc{vi}(\\hat{\\bm{c}}_\\textsc{map},\\hat{\\bm{c}}_{1-\\alpha,\\textsc{map}})$\n -- ------------ --------------------------------------------------------------------- ---------------------------------------------------------------------------------\n 0.75 0.0092 0.5251\n 1.5 0.0180 1.4215\n 3 0.0414 2.2243\n\n : Summary of results for the first simulation study. ", "Data points are generated from , for various values of $\\sigma^2$. \\[tab1\\]\n\nIn Figure \\[fig:sim1\\_co\\] we graphically compare the co-clustering matrices $\\bm{S}$ and $\\bm{S}_\\textsc{oracle}$, which are almost indistinguishable. ", "A more precise quantification of the discrepancy between these two quantities is given in Table \\[tab1\\], where we report the average absolute deviations between the entries of $\\bm{S}$ and $\\bm{S}_\\textsc{oracle}$, for a few values of $\\sigma^2$. Although the differences increase slightly when $\\sigma^2$ is large, this simulation suggests that Gibbs posteriors may be very close to the oracle distribution. ", "In Figure \\[fig:sim1\\_miscl\\] we show the misclassification probabilities for the oracle and for our <span style=\"font-variant:small-caps;\">gb-ppm</span> model which are, unsurprisingly, almost identical.", "\n\nRobust clustering {#sec:sim2}\n-----------------\n\nIn the second simulation study we also consider $n = 200$ observations evenly divided in $K = 4$ clusters, each having $n_1 = \\cdots = n_4 = 50$ data points. ", "Within each partition we assume that $$\\label{sim2}\n(\\bm{x}_i \\mid \\bm{\\mu}_k, \\sigma^2, c_i = k) \\overset{\\textup{iid}}{\\sim} t_2\\left(\\bm{\\mu}_k, \\sigma^2 I_2\\right), \\qquad i \\in C_k, \\quad k=1,\\dots,K,$$ where $t_2(\\bm{\\mu}, \\bm{\\Sigma})$ is a multivariate Student’s $t$-distribution with location $\\bm{\\mu}$, scale $\\bm{\\Sigma}$, and $2$ degrees of freedom. ", "We let $\\bm{\\mu}_1 = (-2,-2)$, $\\bm{\\mu}_2 = (-2,2)$, $\\bm{\\mu}_3 = (2,-2)$, $\\bm{\\mu}_4 = (2,2)$ and $\\sigma^2 = 1$. The random vectors $\\bm{x}_i$ in equation  have finite expectation but infinite variance, meaning that “outliers” are expected. ", "We consider two different <span style=\"font-variant:small-caps;\">gb-ppm</span>s, namely the k-means model of Section \\[sec:squared\\_euclidean\\] and the <span style=\"font-variant:small-caps;\">gb-ppm</span> with Manhattan dissimilarities, which arises when $p=1$ in the class described in Section \\[sec:Minkowski\\]. ", "For uncertainty quantification we estimate $\\lambda$ using the default strategies discussed in the aforementioned sections. ", "Neither model is compatible with the generative process in equation , but this is not a concern within our generalized Bayes framework.", "\n\n![", "Average silhouette width associated to $\\hat{\\bm{c}}_\\textsc{map}$ for different values of $K$, using the k-means algorithm (left plot) and the k-dissimilarities algorithm with Manhattan distances (right plot). ", "Data points are generated according to equation . ", "\\[fig:sim2\\_select\\]](Sim2_select){width=\"\\textwidth\"}\n\nPoint estimates $\\hat{\\bm{c}}_\\textsc{map}$ are obtained in both cases through the k-dissimilarity algorithm. ", "Consistent with the discussion of Section \\[sec:bissiri\\], we select the number of clusters $K$ in an exploratory phase before uncertainty quantification. ", "In practice, one can rely on well-known heuristic methods; eg refer to  @Kaufman1990 [@Hastie2008]. ", "We use the average silhouette statistic [@Kaufman1990], which is a widely used goodness-of-fit index for dissimilarity matrices. ", "The results are reported in Figure \\[fig:sim2\\_select\\], which shows that the model with Manhattan dissimilarities favors $K = 4$ clusters, whereas this criteria in the squared Euclidean case leads to $K = 7$ clusters. ", "In practice, one may prefer the Manhattan model because it leads to a simpler clustering solution. ", "This marked difference in the silhouette statistics is likely due to the presence of outliers, which will be clear from our quantification of the uncertainty.", "\n\n![", "Co-clustering probabilities for the oracle model, the squared Euclidean <span style=\"font-variant:small-caps;\">gb-ppm</span>, and the <span style=\"font-variant:small-caps;\">gb-ppm</span> with Manhattan dissimilarities. ", "Colors correspond to probabilities, from white (low probability) to dark blue (high probability). ", "Data points are generated from equation  and ordered according to the true partition $\\bm{c}_0$. \\[fig:sim2\\_co\\]](Sim2_Co){width=\"\\textwidth\"}\n\nTo make the results comparable, we fix $K = 4$ in both <span style=\"font-variant:small-caps;\">gb-ppm</span>s and run Gibbs samplers for $5,000$ iterations, having discarded $1,000$ samples as burn-in. ", "The traceplots of the losses show good mixing and no evidence against convergence. ", "We compute the associated co-clustering matrices and the oracle matrix $\\bm{S}_\\textsc{oracle}$, which is defined as in Section \\[sec:sim1\\], having replaced the multivariate Gaussian with the multivariate Student’s t-distribution. ", "These matrices are displayed in Figure \\[fig:sim2\\_co\\]. ", "This graph strongly suggests that the <span style=\"font-variant:small-caps;\">gb-ppm</span> based on Manhattan pairwise dissimilarities outperforms the one relying on the squared Euclidean loss. ", "Indeed, the co-clustering matrix of the former is highly similar to $\\bm{S}_\\textsc{oracle}$, which is the gold standard. ", "In the k-means case, the presence of outliers leads to highly unreliable uncertainty quantification. ", "These results are not surprising as absolute deviations have long been used in place of squared losses to robustify clustering; this motivated the <span style=\"font-variant:small-caps;\">pam</span> algorithm [@Kaufman1990], which is closely related to our <span style=\"font-variant:small-caps;\">gb-ppm</span> with pairwise dissimilarities.", "\n\nThe carcinoma dataset {#sec:realdata}\n---------------------\n\nIn the carcinoma dataset [@Agresti2002 Chap. ", "13], seven pathologists separately classified $n=118$ slides regarding the presence or absence of carcinoma of the uterine cervix. ", "We are interested in finding groups among these ratings, in order to identify interpretable patterns. ", "In addition, we aim at providing a probabilistic assessment of the cluster allocation if a new slide were examined.", "\n\n![", "Values of the binary loss $\\ell(\\hat{\\bm{c}}_\\textsc{map}; \\bm{X})$ of equation  evaluated at the <span style=\"font-variant:small-caps;\">map</span>, for the carcinoma dataset and for different values of $K$. \\[fig:app\\_select\\]](App_select){width=\"80.00000%\"}\n\nThe dataset $\\bm{X}$ comprises $118 \\times 7$ binary indicators, implying that the <span style=\"font-variant:small-caps;\">gb-ppm</span> of Section \\[sec:kl\\_model\\] is a reasonable model choice. ", "We first obtained the $\\hat{\\bm{c}}_\\textsc{map}$ point estimate for different values of $K$, using Algorithm \\[Algorithm3\\]. ", "We report in Figure \\[fig:app\\_select\\] the sequence of minimized losses, which decreases as a function of $K$. Application of the elbow rule suggests that between $2$ and $4$ clusters are needed to properly summarize the data. ", "We choose $K = 3$, also on the basis of the following qualitative judgement. ", "Specifically, we expect the evaluations of the pathologist to agree in most cases, meaning that each subject should get either mostly negative or mostly positive diagnoses. ", "These cases represent the two main clusters. ", "However, there might be a third group of subjects whose diagnosis is unclear, which indeed represents a refinement with respect to the initial dichotomic scenario. ", "This description will be empirically confirmed in the data, supporting the choice $K=3$.\n\n![", "Co-clustering probabilities for the Kullback-Leibler <span style=\"font-variant:small-caps;\">gb-ppm</span> applied to the carcinoma dataset. ", "Dashed lines partition the data according to the point estimates $\\hat{\\bm{c}}_\\textsc{map}$ (left plot) and $\\hat{\\bm{c}}_\\textsc{vi}$ (right plot). ", "Observations are ordered in both plots according to the $\\hat{\\bm{c}}_\\textsc{map}$ estimate. ", "\\[fig:app\\_co\\]](App_Co){width=\"\\textwidth\"}\n\nWe run the Gibbs sampler for $16,000$ iterations and we discard the first $1,000$ samples as a burn-in. ", "In the left plot of Figure \\[fig:app\\_co\\] we depict the co-clustering matrix $\\bm{S}$, together with the estimated partition $\\hat{\\bm{c}}_\\textsc{map}$. Although the <span style=\"font-variant:small-caps;\">map</span> is mostly consistent with the pairwise probabilities in $\\bm{S}$, there are six subjects in the second cluster that one would expect to be allocated in the first group according to $\\bm{S}$. This is due to the fact that the <span style=\"font-variant:small-caps;\">map</span> is not necessarily a good point estimate if one is interested in “average” behaviors, such as those of the co-clustering matrix $\\bm{S}$. To overcome this difficulty, we rely on the alternative point estimate $\\hat{\\bm{c}}_\\textsc{vi}$, which is the value minimizing the posterior variation of information distance. ", "As described in @Wade2018, the estimate $\\hat{\\bm{c}}_\\textsc{vi}$ can be approximated by considering a lower-bound of the posterior expectation of $d_\\textsc{vi}(\\bm{c},\\bm{c}')$. Albeit $\\hat{\\bm{c}}_\\textsc{vi}$ is computationally less convenient than $\\hat{\\bm{c}}_\\textsc{map}$, it indeed provides a better posterior summary, as illustrated in Figure \\[fig:app\\_co\\].", "\n\n Pathologist <span style=\"font-variant:small-caps;\">a</span> <span style=\"font-variant:small-caps;\">b</span> <span style=\"font-variant:small-caps;\">c</span> <span style=\"font-variant:small-caps;\">d</span> <span style=\"font-variant:small-caps;\">e</span> <span style=\"font-variant:small-caps;\">f</span> <span style=\"font-variant:small-caps;\">g</span>\n --------------------- ------------------------------------------------- ------------------------------------------------- ------------------------------------------------- ------------------------------------------------- ------------------------------------------------- ------------------------------------------------- -------------------------------------------------\n $\\tilde{\\bm{x}}_1 $ 0.06 0.14 0.01 0.01 0.06 0.01 0.01\n $\\tilde{\\bm{x}}_2 $ 0.56 0.98 0.02 0.06 0.77 0.02 0.65\n $\\tilde{\\bm{x}}_3 $ 0.99 0.97 0.87 0.61 0.99 0.49 0.99\n\n : Adjusted centroids $\\tilde{\\bm{x}}_j = (\\tilde{x}_{j1},\\dots,\\tilde{x}_{j7})^\\intercal$, $j=1,\\dots,3$, of the <span style=\"font-variant:small-caps;\">gb-ppm</span> associated to the variation of information estimate $\\hat{\\bm{c}}_\\textsc{vi}$. \\[tab2\\]\n\n Data points Cluster 1 Cluster 2 Cluster 3\n -------------------------------- ----------- ----------- -----------\n $\\bm{x}_{n+1}=(0,1,0,0,0,0,0)$ 0.79 0.21 0.00\n $\\bm{x}_{n+2}=(0,1,0,0,1,0,0)$ 0.06 0.94 0.00\n $\\bm{x}_{n+3}=(1,1,1,0,1,0,1)$ 0.00 0.04 0.96\n\n : Predictive probabilities for the data points $\\bm{x}_{n+1}$, $\\bm{x}_{n+2}$, and $\\bm{x}_{n+3}$ calculated as in , using the $\\hat{\\bm{c}}_\\textsc{vi}$ point estimate as reference. ", "\\[tab3\\]\n\nIn Table \\[tab2\\] we report the adjusted centroids associated to the point estimate $\\hat{\\bm{c}}_\\textsc{vi}$. Interestingly, they have a clear and simple interpretation. ", "The first cluster refers to cases where all pathologists agree there is no carcinoma, with the occasional exception of $\\textsc{b}$. Conversely, the third cluster refers to cases where most pathologists (<span style=\"font-variant:small-caps;\">a</span>, <span style=\"font-variant:small-caps;\">b</span>, <span style=\"font-variant:small-caps;\">c</span>, <span style=\"font-variant:small-caps;\">e</span>, and <span style=\"font-variant:small-caps;\">g</span>) agree with its presence. ", "Finally, the second cluster refer to cases where there is disagreement among the doctors. ", "This analysis is consistent with the finding of @Agresti2002 and supports the choice $K = 3$. For illustrative purposes, we also report in Table \\[tab3\\] the predictive allocation probabilities associated to $3$ new potential data points. ", "For example, consistently with the above discussion, when the pathologist <span style=\"font-variant:small-caps;\">a</span>, <span style=\"font-variant:small-caps;\">b</span>, <span style=\"font-variant:small-caps;\">c</span>, <span style=\"font-variant:small-caps;\">e</span>, and <span style=\"font-variant:small-caps;\">g</span> agree about the presence of the carciroma, the probability of this new data point being allocated to the third cluster is about $0.96$.\n\nDiscussion {#sec:discussion}\n==========\n\nIn this paper we presented a wide class of generalized Bayes models based on Gibbs posteriors termed <span style=\"font-variant:small-caps;\">gb-ppm</span>. ", "We studied its properties, proposed novel inferential routines and suggested practical usages. ", "Our modeling overcome many limitations of standard Bayesian mixtures, since it leads to computationally efficient algorithms and robust specifications while allowing uncertainty quantification. ", "In addition, our approach merges loss- and model-based approaches into a unified inferential framework. ", "However, the aforementioned advantages are not evenly shared by all <span style=\"font-variant:small-caps;\">gb-ppm</span>s. Indeed, each loss has its own peculiarities, and it favors the formation of specific clusters’ shapes. ", "Thus, the loss function should be carefully selected on the basis of the applications’ aims. ", "Note that our probabilistic interpretation of the Bregman and pairwise dissimilarity cohesions could be very helpful in this sense. ", "This reasoning is consistent with the notion of “optimal” partition $\\bm{c}_\\textsc{opt}$ described in Section \\[sec:bissiri\\], which therefore should not be interpreted as optimal in absolute terms but only in reference to the specific application. ", "Although the sensitivity to the loss’ choice parallels the difficulty in the kernel’s choice in standard mixture models, note that i) generalized Bayes models might have crucial computational advantages, ii) the choice of the loss is often easier and conceptually simpler than the elicitation of a probabilistic model, and iii) losses can be regarded as a generalization of likelihoods, therefore enlarging the modeling possibilities rather than representing an alternative framework. ", "Finally, we note that the choice of the $\\lambda$ parameter remains partially unaddressed beyond the two broad classes of losses described in this paper. ", "Although few general strategies are discussed e.g. in @Bissiri2016 [@Holmes2017], their concrete application to clustering problems is unclear. ", "We believe this issue is an interesting research direction for future works.", "\n\nAcknowledgements {#acknowledgements .unnumbered}\n================\n\nThis work was partially supported by grants R01ES027498 and R01ES028804 of the National Institute of Environmental Health Sciences of the United States National Institutes of Health.", "\n\nComputational details for a <span style=\"font-variant:small-caps;\">gb-ppm</span> with average dissimilarities {#app1}\n=============================================================================================================\n\nIn a <span style=\"font-variant:small-caps;\">gb-ppm</span> with average dissimilarities, a recursive formula for the the differences between the discrepancies is available. ", "The latter appears in Algorithm \\[Algorithm4\\] and in Gibbs sampling via Theorem \\[GibbsSampling\\]. ", "This leads to considerable computational improvements, because one does not need to re-compute the discrepancies $\\mathcal{D}_\\gamma(\\bm{x}_{i'}; \\bm{X}_k)$ at every step of the algorithms. ", "Specifically, let $\\mathcal{D}_\\gamma(\\bm{x}_i; \\bm{X}_k) = 1/n_k \\sum_{i' \\in C_k} \\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p)$ be the average dissimilarity as in equation . ", "Then, $$\\sum_{i' \\in C_k} \\mathcal{D}_\\gamma(\\bm{x}_{i'}; \\bm{X}_k) - \\sum_{i' \\in C_{k,-i}} \\mathcal{D}_\\gamma(\\bm{x}_{i'}; \\bm{X}_{k,-i}) = \\frac{2}{n_k}\\sum_{i' \\in C_k} \\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p) - \\frac{1}{n_k}\\sum_{i' \\in C_{k,-i}}\\mathcal{D}_\\gamma(\\bm{x}_{i'}; \\bm{X}_{k,-i}).$$ Hence, at each step one can update the real numbers $\\sum_{i' \\in C_k} \\mathcal{D}_\\gamma(\\bm{x}_{i'}; \\bm{X}_k)$, $\\sum_{i' \\in C_{k,-i}} \\mathcal{D}_\\gamma(\\bm{x}_{i'}; \\bm{X}_{k,-i})$ by simply adjusting for the factor $2/n_k\\sum_{i' \\in C_k} \\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p)$, which is easy to compute.", "\n\nProofs\n======\n\nProof of Theorem \\[GibbsSampling\\] {#proof-of-theoremgibbssampling .unnumbered}\n----------------------------------\n\nBy definition, the desired full conditional distribution is equal to $\\pi(c_i \\mid \\bm{c}_{-i},\\bm{X}) \\propto \\pi(\\bm{c} \\mid \\lambda, \\bm{X})$. Therefore for any $\\bm{c} : |\\bm{C}| = K$, the probability of $c_i$ being allocated to one of the previous values is $$\\begin{aligned}\n\\mathds{P}(c_i = k \\mid \\bm{c}_{-i}, \\lambda, \\bm{X}) &\\propto \\prod_{j\\neq k} \\rho(C_{j,-i};\\lambda, \\bm{X}_{j,-i}) \\rho(C_k,\\lambda, \\bm{X}_k), \\qquad k=1, \\dots, K. \\\\\n\\end{aligned}$$ Then, multiply the above term by $1/\\prod_{k=1}^{K} \\rho(C_{k,-i};\\lambda, \\bm{X}_{k,-i})$, so that $$\\begin{aligned}\n\\mathds{P}(c_i = k \\mid \\bm{c}_{-i}, \\lambda, \\bm{X}) & \\propto \\frac{\\rho(C_k,\\lambda, \\bm{X}_k)}{\\rho(C_{k,-i},\\lambda, \\bm{X}_{k,-i})}, \\qquad k=1, \\dots,K,\n\\end{aligned}$$ and the result follows.", "\n\nProof of Theorem \\[profile\\_lik\\] {#proof-of-theoremprofile_lik .unnumbered}\n---------------------------------\n\nWe begin by stating the following Lemma, whose proof has been omitted since it can be found in @Banerjee2005 in the $\\lambda = 1$ case. ", "The extension to a general $\\lambda$ is trivial.", "\n\nLet $\\pi_{\\textsc{ed}}(\\bm{x} \\mid \\bm{\\theta},\\lambda) = \\pi(\\bm{x} \\mid \\lambda) \\exp\\{\\lambda [\\bm{\\theta}^\\intercal\\bm{x} - \\kappa(\\bm{\\theta})]\\}$ belong to an exponential dispersion family. ", "Then there exists a function $\\varphi(\\bm{x})$ such that $$\\lambda [\\bm{\\theta}^\\intercal\\bm{x} - \\kappa(\\bm{\\theta})] = - \\lambda \\mathcal{D}_\\varphi(\\bm{x}; \\mu(\\bm{\\theta})) + \\lambda \\varphi(\\bm{x}),$$ where $\\mu(\\bm{\\theta}) = \\int_\\mathds{X} \\bm{x} e^{\\lambda [\\bm{\\theta}^\\intercal\\bm{x} - \\kappa(\\bm{\\theta})]}\\Pi_\\lambda({ \\,\\textup d}\\bm{x})$.\n\nTherefore, the posterior distribution in a mixture model with exponential dispersion components can be written as $$\\begin{aligned}\n\\pi_\\textsc{ed}(\\bm{c} \\mid \\lambda, \\bm{\\theta}_1,\\dots,\\bm{\\theta}_K, \\bm{X}) &\\propto \\prod_{k=1}^K \\prod_{i \\in C_k} \\pi_{\\textsc{ed}}(\\bm{x}_i \\mid \\bm{\\theta}_k, \\lambda) \\\\\n&\\propto \\prod_{k=1}^K \\prod_{i \\in C_k} \\pi(\\bm{x}_i \\mid \\lambda) \\exp\\left\\{\\lambda [\\bm{\\theta}_k^\\intercal\\bm{x}_i - \\kappa(\\bm{\\theta}_k)]\\right\\} \\\\\n&\\propto \\prod_{k=1}^K \\left[\\prod_{i \\in C_k} \\pi(\\bm{x}_i \\mid \\lambda) \\exp\\{\\lambda \\varphi(\\bm{x}_i)\\}\\right] \\exp\\left\\{- \\lambda\\sum_{i \\in C_k} \\mathcal{D}_\\varphi[\\bm{x}_i; \\mu(\\bm{\\theta}_k)]\\right\\} \\\\\n&\\propto \\prod_{k=1}^K \\exp\\left\\{- \\lambda\\sum_{i \\in C_k} \\mathcal{D}_\\varphi[\\bm{x}_i; \\mu(\\bm{\\theta}_k)]\\right\\},\n\\end{aligned}$$ which holds for any value of $\\bm{\\theta}_1,\\dots,\\bm{\\theta}_K$. Then, the maximum likelihood estimate for each $\\bm{\\theta}_k$ can be computed separately, since the likelihood factorizes. ", "Thus, we seek $$\\hat{\\bm{\\theta}}_k = \\arg \\max_{\\bm{\\theta}} \\prod_{i \\in C_k} \\pi_{\\textsc{ed}}(\\bm{x}_i \\mid \\bm{\\theta}, \\lambda) = \\arg \\max_{\\bm{\\theta}} \\sum_{i \\in C_k} \\left[ \\bm{\\theta}^\\intercal\\bm{x}_i - \\kappa(\\bm{\\theta})\\right], \\qquad k=1,\\dots,K,$$ if they exist. ", "By differentiating the right-hand side of the above equation and equating to zero, we obtain that $\\hat{\\bm{\\theta}}_k$ is the solution of $$\\bar{\\bm{x}}_k = \\nabla \\kappa(\\hat{\\bm{\\theta}}_k) = \\mu(\\hat{\\bm{\\theta}}_k),$$ where the last equality follows from the property of exponential dispersion families [@Jorgensen1987]. ", "Therefore, we have $\\hat{\\bm{\\theta}}_k = \\mu^{-1}(\\bar{\\bm{x}}_k) = \\theta(\\bar{\\bm{x}}_k)$, which concludes the proof.", "\n\nProof of Proposition \\[monotone\\] {#proof-of-propositionmonotone .unnumbered}\n---------------------------------\n\nBroadly speaking, the proof follows because we alternate between two minimization procedures. ", "Let us define $$\\ell(\\bm{c}, \\bm{\\theta}_1,\\dots,\\bm{\\theta}_K; \\bm{X}) = \\sum_{k=1}^K\\left[ -\\log \\pi(\\bm{\\theta}_k \\mid \\lambda) + \\lambda \\sum_{i \\in C_k} \\mathcal{D}_\\varphi(\\bm{x}_i; \\mu(\\bm{\\theta}_k))\\right],$$ and recall that $$\\ell(\\bm{c}; \\bm{X}) = \\min_{\\bm{\\theta}_1,\\dots,\\bm{\\theta}_K} \\ell(\\bm{c}, \\bm{\\theta}_1,\\dots,\\bm{\\theta}_K; \\bm{X}).$$ Let $\\bm{c}^{(t)}$ be the partition at the $t$th iteration and let $\\bm{m}_1^{(t)},\\dots,\\bm{m}_K^{(t)}$ be the corresponding centroids. ", "Then, $$\\begin{aligned}\n \\ell(\\bm{c}^{(t)}; \\bm{X}) = \\ell(\\bm{c}^{(t)}, \\bm{m}_1^{(t)},\\dots,\\bm{m}_K^{(t)}; \\bm{X}) & \\ge \\ell(\\bm{c}^{(t+1)}, \\bm{m}_1^{(t)},\\dots,\\bm{m}_K^{(t)}; \\bm{X}) \\\\\n & \\ge \\ell(\\bm{c}^{(t+1)}, \\bm{m}_1^{(t+1)},\\dots,\\bm{m}_K^{(t+1)}; \\bm{X}) = \\ell(\\bm{c}^{(t+1)}; \\bm{X}).", "\n\\end{aligned}$$ The first inequality follows because the re-allocation step minimizes the associated Bregman divergences, whereas the second follows because the adjusted centroids are obtained as the solution of a minimization problem. ", "Moreover, the partition space is finite and this, together with monotonicity, implies that the number of steps will be finite.", "\n\nProof of Proposition \\[monotone2\\] {#proof-of-propositionmonotone2 .unnumbered}\n----------------------------------\n\nBroadly speaking, the proof follows because Algorithm \\[Algorithm4\\] iterates over $n$ minimization procedures. ", "More precisely, let $\\bm{c}^{(t)} = (c_1^{(t)},\\dots,c_n^{(t)})$ be the partition at the $t$th iteration. ", "Then, $$\\begin{aligned}\n \\ell(\\bm{c}^{(t)}; \\bm{X}) = \\ell(c_1^{(t)},\\dots,c_n^{(t)}; \\bm{X}) & \\ge \\ell(c_1^{(t+1)},\\dots,c_n^{(t)}; \\bm{X}) \\ge \\cdots \\ge \\ell(c_1^{(t+1)},\\dots,c_n^{(t+1)}; \\bm{X}) = \\ell(\\bm{c}^{(t+1)}; \\bm{X}).", "\n\\end{aligned}$$ Moreover, the partition space is finite and this, together with monotonicity, implies that the number of steps will be finite.", "\n\nProof of Theorem \\[pairwise\\_lik\\] {#proof-of-theorempairwise_lik .unnumbered}\n----------------------------------\n\nThe proof of Theorem \\[pairwise\\_lik\\] follows from the properties of $L^p$ spherical distributions. ", "Indeed, any $L^p$ spherical distribution admits a useful stochastic representation. ", "More precisely, a random vector $\\bm{x} \\sim \\pi_\\textsc{sp}(\\bm{x})$ is $L^p$ spherical if and only if it can be written as $\\bm{x} \\overset{\\text{d}}{=} r \\bm{u}$, where $r > 0$ is a positive random variable which is independent of the random vector $\\bm{u} \\in \\mathds{R}^d$, which follows the $L^p$ uniform distribution [@Gupta1997]. ", "The random variable $r \\sim \\pi_\\textsc{r}(r)$ is the *radius*, because if $\\bm{x} \\sim \\pi_\\textsc{sp}(\\bm{x})$ and $\\bm{x} \\overset{\\text{d}}{=} r \\bm{u}$, then $r \\overset{\\text{d}}{=} ||\\bm{x}||_p$. In addition, the density of the radius $\\pi_\\textsc{r}(r)$ characterizes the density of the whole vector $\\pi_\\textsc{sp}(\\bm{x})$ and vice versa, namely $$\\label{Lp_spherical}\n\\pi_\\textsc{sp}(\\bm{x}) = g(||\\bm{x}||_p^p)= b_{d,p} \\: \\pi_\\textsc{r}(||\\bm{x}||_p)||\\bm{x}||_p^{1-d}, \\qquad \\pi_\\textsc{r}(r) = \\frac{1}{b_{d,p}} r^{d-1} g(r^p),$$ for any $r > 0$ and $\\bm{x} \\in \\mathds{R}^d$, where $b_{d,p} = p^{d-1} \\Gamma(d/p)/[2 \\Gamma(1/p)]^d$ is a normalizing constant with $\\Gamma(\\cdot)$ the gamma function; see @Gupta1997 for details. ", "Therefore, the term $\\exp\\left\\{ -\\lambda/2\\gamma(||\\bm{x}_i - \\bm{x}_{i'}||_p^p)\\right\\}$, appearing in Definition \\[pairwise\\], is proportional to a proper density of an $L^p$ spherical distribution in $\\mathds{R}^d$ if and only if the law of the radius $\\pi_\\textsc{r}$ is well defined. ", "This is equivalent to the required condition $\\int_{\\mathds{R}_+} r^{d-1} \\exp\\left\\{-\\lambda/2 \\gamma(r^p)\\right\\}{ \\,\\textup d}r < \\infty$.\n" ]

{ "pile_set_name": "ArXiv" }

[ 0, 0, 0, 0.015384615384615385, 0, 0.0053475935828877, 0, 0.004694835680751174, 0.012738853503184714, 0.014184397163120567, 0, 0, 0, 0, 0, 0.011976047904191617, 0, 0.018518518518518517, 0.01098901098901099, 0, 0, 0, 0.011494252873563218, 0.003424657534246575, 0, 0, 0.014925373134328358, 0.027522935779816515, 0.02, 0.008695652173913044, 0.00749063670411985, 0.006024096385542169, 0, 0.0064516129032258064, 0.014492753623188406, 0.009174311926605505, 0.009174311926605505, 0.00909090909090909, 0.008547008547008548, 0, 0, 0, 0, 0.005780346820809248, 0, 0, 0, 0, 0, 0, 0.011695906432748537, 0, 0.006779661016949152, 0, 0, 0, 0.004479283314669653, 0.004366812227074236, 0, 0.017543859649122806, 0.0037243947858473, 0, 0.022222222222222223, 0, 0, 0.015719467956469165, 0.002145922746781116, 0, 0, 0.006006006006006006, 0, 0, 0, 0, 0.005442176870748299, 0, 0, 0.004739336492890996, 0, 0.006134969325153374, 0, 0.015151515151515152, 0.005571030640668524, 0, 0.0072992700729927005, 0, 0.017391304347826087, 0.005934718100890208, 0, 0.018518518518518517, 0.002506265664160401, 0, 0.005333333333333333, 0.006944444444444444, 0.0071090047393364926, 0, 0.01694915254237288, 0.005272407732864675, 0.009302325581395349, 0, 0, 0, 0, 0, 0, 0.013513513513513514, 0.008771929824561403, 0.009111617312072893, 0, 0, 0.010830324909747292, 0, 0, 0.021505376344086023, 0.00916030534351145, 0.008130081300813009, 0, 0, 0.013793103448275862, 0.006097560975609756, 0.006369426751592357, 0.002551020408163265, 0.0037735849056603774, 0.005988023952095809, 0, 0, 0.003587443946188341, 0, 0.016666666666666666, 0.011363636363636364, 0.0030627871362940277, 0, 0.005747126436781609, 0, 0.002785515320334262, 0.0019047619047619048, 0.006369426751592357, 0, 0.010471204188481676, 0.0036101083032490976, 0, 0.009508716323296355, 0.0014925373134328358, 0, 0.01001669449081803, 0, 0.010067114093959731, 0, 0.005494505494505495, 0.0034482758620689655, 0, 0.006060606060606061, 0.018867924528301886, 0, 0.007017543859649123, 0.0035587188612099642, 0, 0, 0, 0, 0.00966183574879227, 0.001984126984126984, 0, 0.006550218340611353, 0, 0.004024144869215292, 0.0032258064516129032, 0, 0, 0.006514657980456026, 0.010101010101010102, 0.0029895366218236174, 0, 0, 0.007543103448275862, 0, 0, 0.008695652173913044, 0, 0, 0, 0.008108108108108109, 0, 0, 0.014925373134328358, 0.0035714285714285713, 0.0072992700729927005, 0.00390625, 0, 0.007125890736342043, 0, 0, 0.010869565217391304, 0.00796812749003984, 0.006493506493506494, 0, 0.006896551724137931, 0.00904977375565611, 0.008086253369272238, 0, 0.004149377593360996, 0.02072538860103627, 0, 0.0043859649122807015, 0, 0, 0.00398406374501992, 0, 0.00411522633744856, 0, 0.009124087591240875, 0.0068143100511073255, 0.013986013986013986, 0, 0.011516314779270634, 0.0041753653444676405, 0.0050578034682080926, 0.009174311926605505, 0.006134969325153374, 0, 0.004830917874396135, 0, 0.0034482758620689655, 0, 0.010139416983523447, 0.01092896174863388, 0.004291845493562232, 0.012195121951219513, 0.0045871559633027525, 0.011049723756906077, 0.006172839506172839, 0.004761904761904762, 0, 0.004484304932735426, 0.008264462809917356, 0.0111731843575419, 0, 0, 0.00851581508515815, 0.010526315789473684, 0, 0.007874015748031496, 0.006622516556291391, 0.005390835579514825, 0.013157894736842105, 0.004166666666666667, 0, 0.005025125628140704, 0, 0.005780346820809248, 0.010344827586206896, 0, 0.015384615384615385, 0.008595988538681949, 0, 0, 0, 0, 0.000946073793755913, 0.004975124378109453, 0, 0, 0, 0, 0.0038314176245210726, 0, 0, 0.002109704641350211, 0, 0, 0, 0, 0.004878048780487805, 0, 0.009569377990430622, 0.011019283746556474, 0.0040650406504065045, 0.006369426751592357, 0, 0.007407407407407408, 0, 0, 0, 0.006024096385542169, 0.0064516129032258064, 0.02, 0.007751937984496124, 0, 0, 0, 0, 0, 0, 0.005780346820809248, 0, 0.004310344827586207, 0, 0, 0, 0, 0.0029585798816568047, 0.027777777777777776, 0, 0, 0, 0, 0, 0.007936507936507936, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.013333333333333334, 0, 0.002688172043010753, 0, 0.01098901098901099, 0, 0, 0.008368200836820083, 0.004580152671755725, 0, 0, 0, 0, 0, 0, 0.004, 0.002061855670103093, 0, 0.013888888888888888, 0, 0.00796812749003984, 0, 0, 0, 0, 0.006546644844517185, 0.010893246187363835, 0.008, 0, 0.015151515151515152, 0.009502923976608187, 0.0070921985815602835, 0.003067484662576687, 0.008333333333333333, 0, 0.008064516129032258, 0.013245033112582781, 0, 0, 0.004347826086956522, 0.009433962264150943, 0.008547008547008548, 0, 0, 0, 0.008875739644970414, 0.00938337801608579, 0, 0.007042253521126761 ]

[ "Tim Broomhead\n\nTim Broomhead (born 22 March 1994) is a professional Australian rules football player who plays for Collingwood in the Australian Football League (AFL). ", "He was recruited by the club with draft pick #20 in the 2012 AFL Draft.", "\n\nDue to getting glandular fever during the 2013 NAB Cup, Broomhead had to wait until Round 13 in 2014 to make his debut, coming on as a substitute against at Etihad Stadium. ", "Broomhead remained in the side for the following game against Hawthorn and started as the substitute again but came on early during the second quarter replacing Tom Langdon with concussion. ", "The following week he was dropped and returned to the VFL.", "\n\nIn round 2, 2018 against at the Melbourne Cricket Ground, Broomhead suffered a broken leg when he kicked the goal post after attempting to soccer the ball through. ", "At the conclusion of the 2018 season, Collingwood delisted Broomhead, stating they will re-draft him as a rookie if he is available. ", "A month later, Collingwood re-drafted Broomhead as a rookie with pick 17.", "\n\nStatistics\n Statistics are correct to the end of the 2018 season\n\n|- style=\"background:#eaeaea;\"\n! ", "scope=\"row\" style=\"text-align:center\" | 2013\n|\n| 29 || 0 || — || — || — || — || — || — || — || — || — || — || — || — || — || —\n|- \n! ", "scope=\"row\" style=\"text-align:center\" | 2014\n|\n| 29 || 8 || 9 || 4 || 60 || 43 || 103 || 21 || 19 || 1.1 || 0.5 || 7.5 || 5.4 || 12.9 || 2.6 || 2.4\n|- style=\"background:#eaeaea;\"\n! ", "scope=\"row\" style=\"text-align:center\" | 2015\n|\n| 29 || 11 || 7 || 5 || 81 || 79 || 160 || 16 || 39 || 0.6 || 0.5 || 7.4 || 7.2 || 14.5 || 1.5 || 3.5\n|- \n! ", "scope=\"row\" style=\"text-align:center\" | 2016\n|\n| 29 || 2 || 3 || 0 || 5 || 11 || 16 || 3 || 3 || 1.5 || 0 || 2.5 || 5.5 || 8.0 || 1.5 || 1.5\n|- style=\"background:#eaeaea;\"\n! ", "scope=\"row\" style=\"text-align:center\" | 2017\n|\n| 29 || 14 || 8 || 11 || 100 || 129 || 229 || 47 || 35 || 0.6 || 0.8 || 7.1 || 9.2 || 16.4 || 3.4 || 2.5\n|- \n! ", "scope=\"row\" style=\"text-align:center\" | 2018\n|\n| 29 || 1 || 0 || 0 || 0 || 0 || 0 || 0 || 0 || 0.0 || 0.0 || 0.0 || 0.0 || 0.0 || 0.0 || 0.0\n|- class=\"sortbottom\"\n! ", "colspan=3| Career\n! ", "36\n! ", "27\n! ", "20\n! ", "246\n! ", "262\n! ", "508\n! ", "87\n! ", "96\n! ", "0.8\n! ", "0.6\n! ", "6.8\n! ", "7.3\n! ", "14.1\n! ", "2.4\n! ", "2.7\n|}\n\nReferences\n\nExternal links\n\nCategory:1994 births\nCategory:Living people\nCategory:Collingwood Football Club players\nCategory:Australian rules footballers from South Australia\nCategory:Port Adelaide Football Club (SANFL) players\nCategory:Port Adelaide Football Club players (all competitions)" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0.02976190476190476, 0, 0, 0.005263157894736842, 0.017241379310344827, 0.005988023952095809, 0.015037593984962405, 0.0136986301369863, 0.009900990099009901, 0, 0.0055248618784530384, 0, 0.005747126436781609, 0, 0, 0.05, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.013422818791946308 ]

[ "Tutorial: Using MIDI to Control Animation in After Effects | Part 1\n\nBy Joey Korenman\n\nHere's how to control animation with a MIDI controller.", "\n\nReady to test the limits of what's possible in After Effects? ", "In this lesson you're going to unleash some real creativity with using a piece of software for something that it wasn't really built for in the first place. ", "There's going to be some plugins and expressions that you'll need for this lesson, so head on over to the resources tab to grab those.", "\n\nDownload the MIDI controlled animation project file for After Effects." ]

{ "pile_set_name": "Pile-CC" }

[ 0.02112676056338028, 0, 0, 0, 0 ]

[ "Percutaneous ethanol sclerotherapy of autonomous thyroid nodules: preliminary results.", "\nPercutaneous ethanol sclerotherapy was applied to 20 patients (17 women and 3 men, aged 30 to 77 yrs) with autonomous thyroid nodules. ", "The patients either were contraindicated because of hazardous concomitant diseases or refused surgical or 131I treatment. ", "Ethanol (95 alc%) was injected in 4 to 6 sessions at a mean dose of 0.63 ml/cc nodular tissue. ", "Physical examination, T3, T4 and TSH assays, thyroid scintigraphy, sonography, and fine needle aspiration biopsy were used to evaluate the treatment results within a 12-month follow-up. ", "Two months after ethanol injection all patients showed an improvement; in six patients the symptoms recurred in the 6th month. ", "Two of them underwent a second cycle of ethanol injection. ", "12 months later an euthyroid state was maintained in 16 (80%) patients, while 4 (20%) did not respond to the treatment. ", "The treatment results were markedly determined by the nodule size as no effect was achieved in nodules larger than 15-20 ml in volume. ", "This method appears to be an appropriate and harmless alternative to the surgical and 131I treatment of autonomous thyroid nodules less than 15-20 ml in volume, especially in patients with grave concomitant diseases." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0.010526315789473684, 0, 0, 0, 0, 0, 0 ]

[ "By Strange Mysteries\n\nWas earth visited in the past by Ancient Aliens flying UFOs around and telling us where to build their cities….perhaps hoping that one day in the future we would put together the clues?", "\n\nAn uncanny and what seems to be a planned alignment of a multitude of ancient archaeological sites around Earth strongly points towards humans of the past being guided globally and over many millennia by some sort of higher intelligence. ", "Personally, my mind was absolutely blown once I manually connected all the sites on Google Earth myself, which is exactly what I’m about to do for you. ", "At the end of the video, I expect your mind to be blown just as mine was. ", "I also encourage you to try this out for yourself in Google Earth and replicate my results.", "\n\nBefore we begin, please forgive me for butchering some of these names.", "\n\nNearly all of these sites are considered what is known as a UNESCO World Heritage Site, sites that are of extreme special cultural or physical significance.", "\n\n1) Easter Island:\n\nWe’re going to start off with Easter Island, which is home to the mysterious Moai statues which were created around 800 years ago for reasons that still remain unknown. ", "A total of 887 monolithic stone Moai statues have been counted.", "\n\n2) Nazca Lines:\n\nThe Nazca Lines are mysterious giant glyphs carved into the ground in the Nazca Desert of Peru. ", "They were thought to be built around 1600 years ago – no one knows why they are there as they can only be seen from the sky…why would anyone do that when flight hadn’t been invented yet?", "\n\n3) Machu Picchu:\n\nMachu Piccu was built around 600 years ago at the height of the Incan Empire and allegedly as a means to honor a “sacred landscape.” ", "If you haven’t heard of it before, it’s a city literally built on top of a mountain by hand….with rocks sometimes weighing as much as 40 tons.", "\n\n4) Tassili n’Ajjer\n\nTassili n’Ajjer is a range of mountains in Algeria in the Sahara Desert. ", "It contains some 15,000 ancient cave drawings some of which are up to 10,000 years old. ", "Some of the carvings are a bit strange as well.", "\n\n5) Persepolis\n\nPersepolis is the ancient capital of Persia. ", "It was built around 2500 years ago.", "\n\n6) Khajuraho Group of Monuments\n\nThe Khajuraho Group of Monuments are 85 special temples to honor Hindu gods built around 1000 years ago. ", "They were allegedly built in this area as it is an area where quote “the gods love to play” according to local legend. ", "The temples are dedicated to various Hindu gods such as Shiva, Vishnu, Ganesha, and Jain Tithanks.", "\n\n7) The Oasis at Siwa\n\nThe Oasis at Siwa is estimated to be around 4,600 years old..and many believe it could be much older. ", "It is famous for being the home of the Oracle of Amun. ", "Amun was an Egyptian Diety. ", "Alexander the Great believed that his divinity was a result of his visit to this oracle.", "\n\n8) The Pyramids of Giza\n\nI’m sure you’ve heard of the Pyramids of Giza before. ", "They were estimated to have been constructed around 4,600 years ago.", "\n\n9) Mohenjo Daro\n\nMohenjo Daro was built in around 4600 years ago. ", "It is part of the ancient Indus Valley Civilization. ", "The name literally translates to the Mound of the Dead. ", "It is thought to be one of the world’s earliest major urban settlements.", "\n\n10) Angkor Wat\n\nAngkor Wat was built around 1000 years ago and was dedicated to the god Vishnu. ", "According to a 13th Century Chinese traveler, it was allegedly believed that the temple was contructed in only one night by an architect of divine power.", "\n\nNow let’s complete our journey by connecting the final destination to our original one, Easter Island.", "\n\nOnce we zoom out….do you see a pattern? ", "ALL of these sites line up on the same line around the Earth. ", "The line is, except for a few very minor deviations that may be due to the Earth slightly shifting over the many millenia these sites were constructed as well as perhaps partial user error on my part mapping these out…..perfect. ", "It almost looks like an equator. ", "The probability of this occuring by chance is basically zero.", "\n\nWhy do all these ancient and quite important sites line up on the same line on Earth? ", "Was someone guiding the construction of all these ancient sites? ", "Why are so many if not all of the sites related to something divine….and what was that divinity?", "\n\nIn order to orchestrate such a thing, these civilizations would have somehow had to have made contact with each other across the space of the entire planet as well as over the course of several millenia in addition to being able to basically see the Earth from space in order to create such an uncanny alignment.", "\n\nSource : Strange Mysteries\n\nIf you found this article interesting please share it across social media.", "\n\nAlso, why not check out our Instagram page here." ]

{ "pile_set_name": "OpenWebText2" }

[ 0.004830917874396135, 0, 0, 0, 0, 0, 0.006329113924050633, 0, 0, 0.008695652173913044, 0, 0.013071895424836602, 0, 0, 0, 0, 0, 0, 0.007142857142857143, 0, 0.02040816326530612, 0, 0.01818181818181818, 0.03571428571428571, 0.011363636363636364, 0.012345679012345678, 0, 0.014705882352941176, 0, 0.017857142857142856, 0, 0.01020408163265306, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "Q:\n\nHow can I get a \"flat\" output of data from a parent / child table with LINQ\n\nI have two tables. ", "The Word table contains a list of words and then the WordForm table contains a list of parts of speech for each word.", "\npublic class Word\n{\n public Word()\n {\n CategoryId = 1;\n WordForms = new System.", "Collections.", "Generic.", "List<WordForm>();\n }\n public string WordId { get; set; } // WordId (Primary key) (length: 20)\n public int CategoryId { get; set; } // CategoryId\n public virtual System.", "Collections.", "Generic.", "ICollection<WordForm> WordForms { get; set; } // WordForm.", "FK_WordFormWord\n public virtual WordCategory WordCategory { get; set; } // FK_WordWordCategory\n\n}\n\npublic class WordForm\n{\n public string WordFormId { get; set; } // WordFormId (Primary key) (length: 20)\n public string WordId { get; set; } // WordId (length: 20)\n public int PosId { get; set; } // PosId\n public string Definition { get; set; } // Definition\n public virtual Pos Pos { get; set; } // FK_WordFormPos\n public virtual Word Word { get; set; } // FK_WordFormWord\n}\n\nI tried joining these two tables like this:\nvar word = db.", "Words\n .Include(wf => wf.", "WordForms)\n .AsNoTracking()\n .FirstOrDefaultAsync();\n\nThis gives me an output in object form. ", "\nHow can I get to see what is in the tables so that I have returned to me data like this where each row has all of these columns:\nWordId CategoryId WordFormId PosId Definition\nWordId CategoryId WordFormId PosId Definition\nWordId CategoryId WordFormId PosId Definition\nWordId CategoryId WordFormId PosId Definition\n\nA:\n\nYou can use SelectMany to flatten the output:\n var words = db.", "Words\n .Include(wf => wf.", "WordForms)\n .SelectMany(w => w.WordForms.", "Select(wf => new\n {\n WordId = w.WordId,\n CategoryId = w.CategoryId,\n WordFormId = wf.", "WordFormId,\n PosId = wf.", "PosId,\n Definition = wf.", "Definition\n }));\n\nIf you want the result only for the first Word:\nvar word = await db.", "Words\n .Include(wf => wf.", "WordForms)\n .AsNoTracking()\n .FirstOrDefaultAsync();\n\nvar output = word.", "WordForms.", "Select(wf => new\n{\n WordId = wf.", "Word.", "WordId,\n CategoryId = wf.", "Word.", "CategoryId,\n WordFormId = wf.", "WordFormId,\n PosId = wf.", "PosId,\n Definition = wf.", "Definition\n});\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0.008547008547008548, 0.01, 0, 0.125, 0.016666666666666666, 0, 0.125, 0, 0.005405405405405406, 0, 0.008620689655172414, 0, 0, 0.01818181818181818, 0.0196078431372549, 0, 0, 0, 0, 0.009259259259259259, 0.1, 0, 0, 0.03571428571428571, 0, 0.03125, 0, 0, 0 ]

[ "Illinois Emergency Management Agency Lands Terror Training Money\n\n/ MyRadioLink.com\n\nJuly 14, 2017 07:12 am\n\nIllinois Emergency Management Agency Lands Terror Training Money\n\n(Springfield, IL) — Illinois is getting just over a million dollars to help prepare for the worst. ", "The Illinois Emergency Management Agency yesterday announced it’s received a million-dollar grant for terror training. ", "IEMA Director James Joseph says the money is to be used to prepare for a complex coordinated terrorist attack." ]

{ "pile_set_name": "Pile-CC" }

[ 0.0072992700729927005, 0.008403361344537815, 0.01818181818181818 ]

[ "Kickers and defenses. ", "They are the fantasy equivalent of old TV credits running in front of a freeze-framed photo. ", "The show is over, but there is still some scrolling left to do.", "\n\nIt is common practice by many to wait until the end to draft the two most unpredictable and easily replaceable fantasy positions — for good reason: because they are unpredictable and easily replaceable.", "\n\nCertainly, the Madman endorses exactly this practice. ", "After loading up on three each of running backs and wide receivers, after then addressing quarterback and tight end while sprinkling in at least two more at RB and WR, then you get to the final roster spots: kicker and defense/special teams.", "\n\nWith rare exception, this is the preferred method. ", "In recent years, we have altered our approach. ", "We now prioritize a stable kicker over an unstable DST. ", "For example, the lowest-ranked DST we like is the Cowboys. ", "If they, or our other favorites who are more likely to be already taken (Bears, Jaguars, Texans) aren’t available in the 15th round of a traditional PPR league, then we will opt to go kicker first, then take the DST with the best Week 1 matchup (whichever team is playing the worst projected offense).", "\n\nWhy? ", "Because we are perfectly comfortable with streaming defenses. ", "We can drop a DST and pick up one with a better matchup week to week. ", "We can be prepared to do this all season, though we hold out hope that during the bye weeks a preferred or emerging unit will become available that we then keep the rest of the way.", "\n\nBut, remember when we said “with rare exception”? ", "Well, that’s an important disqualifier. ", "It might not happen often, but if it does, it can help to bump one or both of those picks up a round or two if you know the player(s) you are targeting for deep bench spots will remain available.", "\n\nMaybe you have a deep sleeper no one else is lauding. ", "Maybe it is a handcuff RB no one else would want. ", "Maybe it is a second QB, and there are plenty to choose from still available. ", "If these situations arise, then go ahead and jump the gun on a better DST. ", "Kickers, we’re not as inclined to treat with the same boost. ", "But if, say, Justin Tucker or Greg Zuerlein or Harrison Butker is still on the board, we could forgive such a reach under these specific conditions.", "\n\nThe point being, most credit rolls are still mundane and offer little entertainment value. ", "But sometimes, there is an extra-credits scene, a stinger that emerges from inanity that offers something of value. ", "When that happens, don’t be afraid to pay attention and adjust your plan accordingly. ", "Maybe the after-credits scene has some value." ]

{ "pile_set_name": "OpenWebText2" }

[ 0, 0, 0, 0, 0.017857142857142856, 0.008298755186721992, 0, 0, 0.017857142857142856, 0.01694915254237288, 0.006644518272425249, 0, 0, 0.014285714285714285, 0, 0, 0, 0, 0, 0, 0, 0.013333333333333334, 0.01639344262295082, 0.02027027027027027, 0, 0, 0, 0 ]

[ "Q:\n\nAndroid Universal Image Loader (nostra13): getting nullPointerException\n\nIt looks like it has some problem setting Bitmap, but I don't know why. ", "It's the first time I use this library, so every help is much appreciated! ", "The row where I'm getting error is where I do setImageBitmap(myBitmap).", "\nHere is the code of my adapter:\npublic class CustomGalleryAdapter extends BaseAdapter {\n\nprivate List<String> imageUrls;\nprivate Context context;\n\n// Variabili per Android Universal Image Loader\nprivate ImageLoader imageLoader;\n\npublic CustomGalleryAdapter(Context context, Context appContext, List<String> imageUrls){\n this.imageUrls = imageUrls;\n this.context = context;\n\n // Inizializza Android Universal Image Loader\n imageLoader = ImageLoader.getInstance();\n imageLoader.init(ImageLoaderConfiguration.createDefault(appContext));\n}\n\n@Override\npublic int getCount() {\n return imageUrls.size();\n}\n\n@Override\npublic Object getItem(int position) {\n return \"file://\"+imageUrls.get(position);\n}\n\n@Override\npublic long getItemId(int position) {\n return position;\n}\n\n@Override\npublic View getView(int position, View convertView, ViewGroup parent){\n\n // Classe che contiene l'imageView\n ViewHolder holder;\n\n // Se viene creato e non \"riutilizzato\"\n if (convertView == null) {\n holder = new ViewHolder();\n holder.imageView = new ImageView(context); \n } \n else {\n holder = (ViewHolder) convertView.getTag();\n }\n\n // Decodifica il Bitmap\n Bitmap myBitmap = BitmapFactory.decodeFile(imageUrls.get(position));\n\n // Setta il bitmap nella imageView\n holder.imageView.setImageBitmap(myBitmap);\n holder.imageView.setLayoutParams(new GridView.", "LayoutParams(GridView.", "LayoutParams.", "WRAP_CONTENT, GridView.", "LayoutParams.", "WRAP_CONTENT));\n\n // Ottiene l'URL formattato per la libreria\n String imageUrl = \"file://\"+imageUrls.get(position);\n\n // Carica e visualizza l'immagine\n imageLoader.displayImage(imageUrl, holder.imageView);\n\n return holder.imageView;\n} \n\n// Classe \"wrapper\" per l'imageView\nstatic class ViewHolder {\n ImageView imageView;\n}\n}\n\nAnd here is LogCat:\n05-12 10:55:23.841: E/AndroidRuntime(8691): FATAL EXCEPTION: main\n05-12 10:55:23.841: E/AndroidRuntime(8691): java.lang.", "NullPointerException\n05-12 10:55:23.841: E/AndroidRuntime(8691): at it.romagnoli.andrea.myrehearsals.adapters.", "CustomGalleryAdapter.getView(CustomGalleryAdapter.java:69)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "AbsListView.obtainView(AbsListView.java:2143)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "GridView.makeAndAddView(GridView.java:1341)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "GridView.makeRow(GridView.java:341)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "GridView.fillDown(GridView.java:283)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "GridView.fillFromTop(GridView.java:417)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "GridView.layoutChildren(GridView.java:1229)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "AbsListView.onLayout(AbsListView.java:1994)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "View.layout(View.java:14008)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewGroup.layout(ViewGroup.java:4373)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "LinearLayout.setChildFrame(LinearLayout.java:1663)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "LinearLayout.layoutVertical(LinearLayout.java:1521)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "LinearLayout.onLayout(LinearLayout.java:1434)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "View.layout(View.java:14008)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewGroup.layout(ViewGroup.java:4373)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "FrameLayout.onLayout(FrameLayout.java:448)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "View.layout(View.java:14008)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewGroup.layout(ViewGroup.java:4373)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "LinearLayout.setChildFrame(LinearLayout.java:1663)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "LinearLayout.layoutVertical(LinearLayout.java:1521)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "LinearLayout.onLayout(LinearLayout.java:1434)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "View.layout(View.java:14008)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewGroup.layout(ViewGroup.java:4373)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.widget.", "FrameLayout.onLayout(FrameLayout.java:448)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "View.layout(View.java:14008)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewGroup.layout(ViewGroup.java:4373)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewRootImpl.performLayout(ViewRootImpl.java:1892)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewRootImpl.performTraversals(ViewRootImpl.java:1711)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewRootImpl.doTraversal(ViewRootImpl.java:989)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "ViewRootImpl$TraversalRunnable.run(ViewRootImpl.java:4351)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "Choreographer$CallbackRecord.run(Choreographer.java:749)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "Choreographer.doCallbacks(Choreographer.java:562)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "Choreographer.doFrame(Choreographer.java:532)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.view.", "Choreographer$FrameDisplayEventReceiver.run(Choreographer.java:735)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.os.", "Handler.handleCallback(Handler.java:725)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.os.", "Handler.dispatchMessage(Handler.java:92)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.os.", "Looper.loop(Looper.java:137)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at android.app.", "ActivityThread.main(ActivityThread.java:5041)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at java.lang.reflect.", "Method.invokeNative(Native Method)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at java.lang.reflect.", "Method.invoke(Method.java:511)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at com.android.internal.os.", "ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:793)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at com.android.internal.os.", "ZygoteInit.main(ZygoteInit.java:560)\n05-12 10:55:23.841: E/AndroidRuntime(8691): at dalvik.system.", "NativeStart.main(Native Method)\n\nA:\n\nIf convertView is not null you don't have a ViewHolder to return with getTag()(you don't set the ViewHolder as a tag anywhere in your code on the convertView). ", "Also if you just use an ImageView you really don't need a ViewHolder, basically the ViewHolder will have a reference to an ImageView, then the ViewHolder will be set as a tag on the view returned by the getView() method which is the exact ImageView the holder points to. ", "Simply use an ImageView.", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0.020134228187919462, 0, 0, 0.012014134275618375, 0, 0.07692307692307693, 0.08695652173913043, 0.07692307692307693, 0.006160164271047228, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.008695652173913044, 0, 0, 0, 0.008264462809917356, 0.008771929824561403, 0, 0, 0.009708737864077669, 0, 0, 0, 0.009615384615384616, 0.009433962264150943, 0.007633587786259542, 0, 0.015228426395939087, 0.02214022140221402, 0.041666666666666664, 0 ]

[ "Q:\n\nmysql -field for text\n\nI need a field in my table to store some text. ", "300 characters is the limit for the text.", "\nWhat is the best choice the option: varchar (300) or text?", "\n\nA:\n\nVARCHAR and TEXT columns are very similar. ", "The manual says:\n\nIn most respects, you can regard a BLOB column as a VARBINARY column that can be as large as you like. ", "Similarly, you can regard a TEXT column as a VARCHAR column.", "\n\nHere are a few differences:\n\nTEXT columns are blank-padded at the end for indexing purposes\nTEXT columns cannot have default values\nVARCHAR columns have only 1 byte of overhead if the number of bytes needed for the text is less than 256. (", "The number of bytes needed for a particular text depends on the encoding.) ", "TEXT columns always have 2 bytes of overhead.", "\nAll text in VARCHAR columns counts against the row limit of 65,535 bytes. ", "The text of TEXT columns is stored separately and does not count against this limit.", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0, 0.008264462809917356, 0, 0, 0, 0, 0, 0, 0 ]

[ "(J Am Heart Assoc. ", "2019;8:e014074 DOI: 10.1161/JAHA.119.014074.)31583935\n\nClinical PerspectiveWhat Is New?We have introduced the treatment device concept of the \"biodegradable flow diverters\" for the cerebral aneurysms, which induce a temporal flow‐diverting effect and provide a vascular scaffold for neointimal formation at the neck of the aneurysm until occlusion.", "The in vivo assessment showed that aneurysm occlusion is possible with these devices and displayed a certain feasibility of the device, especially on the preservation of the branching vessel and the extent of neointima formation.", "What Are the Clinical Implications?Although they are in their initial stage of development, biodegradable flow diverters have potential to be a therapeutic option for cerebral aneurysms in the future.", "\n\nIntroduction {#jah34432-sec-0008}\n============\n\nThe use of biodegradable vascular scaffolds has already been introduced in the field of interventional cardiology for the treatment of coronary artery disease in clinical settings. ", "These scaffolds may potentially overcome the drawbacks of permanent coronary stents, such as late‐stent thrombosis, late in‐stent stenosis, and pathophysiological changes in the vessel wall caused by metallic caging of the artery.[1](#jah34432-bib-0001){ref-type=\"ref\"} However, the first‐generation coronary biodegradable scaffolds were associated with a higher thrombosis rate compared with the standard drug‐eluting stents, which is likely attributable to the large strut thickness.[2](#jah34432-bib-0002){ref-type=\"ref\"}, [3](#jah34432-bib-0003){ref-type=\"ref\"}\n\nIn the field of interventional neuroradiology, biodegradable polymers have been used in some detachable coils for aneurysm occlusion to enhance therapeutic efficacy. ", "For example, poly (lactic‐*co*‐glycolic acid) has been used in the Matrix coil detachable system (Stryker, Minneapolis, MN), and poly (glycolic acid) has been used in the Cerecyte coil (DePuy Synthes, Raynham, MA). ", "Polymer degradation was shown to increase tissue regeneration and cell proliferation response within the aneurysm sac, which is expected to decrease the aneurysm recanalization rate.[4](#jah34432-bib-0004){ref-type=\"ref\"}, [5](#jah34432-bib-0005){ref-type=\"ref\"}\n\nAlthough some animal studies investigated the combined use of biodegradable polymers with metallic stents for intracranial use, no study investigated a device that fundamentally consisted of biodegradable materials intended for intracranial use.[6](#jah34432-bib-0006){ref-type=\"ref\"}, [7](#jah34432-bib-0007){ref-type=\"ref\"}, [8](#jah34432-bib-0008){ref-type=\"ref\"} Herein, we present the concept of a biodegradable flow diverter (BDFD), which provides a transient vascular scaffold for neointimal formation and induces a flow‐diverting effect at the aneurysm neck. ", "After the healing period, the degradation of the BDFD leaves the vessel with a healthy endothelium and normal vasomotion and renders it devoid of caging.", "\n\nThe BDFD has several potential advantages against the metallic flow diverter (FD). ", "Because the FD requires a high metal coverage rate for aneurysm occlusion, it is associated with a relatively high risk of ischemic complications. ", "Remote ischemic events may occur \\>1 year after device placement.[9](#jah34432-bib-0009){ref-type=\"ref\"}, [10](#jah34432-bib-0010){ref-type=\"ref\"}, [11](#jah34432-bib-0011){ref-type=\"ref\"} Dual‐antiplatelet therapy with aspirin and clopidogrel for 6 months after implantation is generally used to prevent late ischemic events, in return for the increased risk of hemorrhagic complications. ", "The absence of residual foreign materials theoretically reduces the risk of late‐stent thrombosis and in‐stent stenosis; this may reduce the risk of remote ischemic complications. ", "Besides, BDFDs will be more appropriate for pediatric patients and patients allergic to metals; in the event of treatment failure, there will be more options for salvage therapy after the BDFD absorption.", "\n\nHowever, there are some concerns pertaining to biodegradable stent use in cerebral vessels. ", "During degradation, biodegradable materials tend to induce a mild but persistent inflammatory response around the strut.[12](#jah34432-bib-0012){ref-type=\"ref\"}, [13](#jah34432-bib-0013){ref-type=\"ref\"} Whether the inflammation and consequent cell proliferation response during degradation are tolerable needs to be studied. ", "In addition, because biodegradable polymers gradually lose their mechanical strength, the device may get deformed and fragmented during degradation.", "\n\nHerein, we report a preclinical study of a BDFD for aneurysm treatment. ", "Because this device is still in its early stage of development, we have focused on its feasibility, especially the biological response and morphometric changes in the parent vessel.", "\n\nMaterials and Methods {#jah34432-sec-0009}\n=====================\n\nThe data that support the findings of this study are available from the corresponding author on reasonable request.", "\n\nBDFD Design {#jah34432-sec-0010}\n-----------\n\nA high‐molecular‐weight poly‐[l]{.smallcaps}‐lactic acid polymer (weight‐average molecular weight \\[Mw\\]=185 000 g/mol) was prepared. ", "This polymer was processed into a thin fiber using a melt‐spinning method until the strut thickness became 40 to 45 μm. ", "Then, the poly‐[l]{.smallcaps}‐lactic acid fibers were woven into a braided structure composed of 48 fibers (Figure [1](#jah34432-fig-0001){ref-type=\"fig\"}A and B). ", "Annealing was performed by heating at 120°C for 2 hours. ", "Three radiopaque gold markers were attached at both ends of the device. ", "The designed size of the BDFD was 4.0×10, 4.0×12, and 4.0×15 mm. ", "The BDFD was mounted onto a delivery system compatible with the 6F guiding catheter, and then the system was navigated to the vascular site with a 0.014‐in guidewire. ", "The BDFD was then pushed out from the delivery system with a delivery wire. ", "Similar to the metallic braided stents, the BDFD undergoes self‐expansion during deployment; however, it is not complete for its designed size. ", "Routine angioplasty was performed after the deployment. ", "The device is designed for degradation, except for the gold markers at both ends.", "\n\n![", "Scaffold design and in vitro evaluation. **", "A**, Photograph of the biodegradable flow diverter used in this study. **", "B**, A scanning electron microscopy image showing 40 to 45 μm braided wires composed of poly‐[l]{.smallcaps}‐lactic acid (bar=400 μm). **", "C**, Results of the accelerated degradation test showing a decrease in the weight‐average molecular weight by 98% at 42 days. ", "Throughout the degradation, crystallinity was maintained at ≈75%. **", "D**, Results of the real‐time degradation test showing degradation of 95% of the device by 1.5 years. ", "Mn indicates number‐average molecular weight; Mw, weight‐average molecular weight.](JAH3-8-e014074-g001){#jah34432-fig-0001}\n\nIn Vitro Accelerated Degradation Test {#jah34432-sec-0011}\n-------------------------------------\n\nTo investigate the in vitro degradation, 10 samples of BDFD were placed in phosphate buffered saline (pH=7.4±0.2) at 70±1°C. ", "On days 0, 3, 7, 10, 12, 14, 21, 28, 35, and 42, one sample was removed from the buffer, rinsed with distilled water, vacuum dried, and weighed. ", "Then, each sample was subjected to gel permeation chromatography and differential scanning calorimetry.", "\n\nThe number‐average molecular weight and Mw were measured using gel permeation chromatography. ", "Gel permeation chromatography was performed at 40°C using a Shimadzu analyzer system. ", "A Tosoh TSKgel SuperHZM‐N column was used with chloroform as the eluent at a flow rate of 0.25 mL/min (Tosoh, Tokyo, Japan). ", "A sample was injected at a concentration of 2.0 mg/mL. Polystyrene standards were used as the reference.", "\n\nFor the differential scanning calorimetry analysis, 2‐mg samples were placed in sealed aluminum pans. ", "The glass transition temperature, melting point of the fibers, and crystallinity were determined by increasing the temperature from 25°C to 220°C at 10°C/min using a thermoregulated differential scanning calorimeter (Bruker AXS, Billerica, MA).", "\n\nIn Vitro Real‐Time Degradation Test {#jah34432-sec-0012}\n-----------------------------------\n\nTwenty‐four samples of the same polymer were placed in phosphate buffered saline (pH=7.4±0.2) at 37±1°C. ", "On days 28 (n= 4), 91 (n=4), 182 (n=6), 273 (n=4), 364 (n=4), and 553 (n=2), number‐average molecular weight and Mw were measured using gel permeation chromatography, similar to the accelerated degradation test.", "\n\nIn Vivo Study Design and Animal Experiment {#jah34432-sec-0013}\n------------------------------------------\n\nAll experimental protocols were approved by the institutional animal care committee. ", "Elastase‐induced aneurysms were created in 18 female New Zealand white rabbits (weight, 2.6--3.0 kg).[14](#jah34432-bib-0014){ref-type=\"ref\"} Anesthesia was induced using an intravenous injection of pentobarbital and was maintained with 1% to 2% isoflurane throughout the procedure. ", "Aneurysms were allowed to mature for at least 21 days before BDFD placement because the aneurysm is reported to mature 2 weeks after aneurysm creation.[14](#jah34432-bib-0014){ref-type=\"ref\"}\n\nAnimals were treated with aspirin (30 mg) for 5 days before BDFD placement. ", "Through a right transfemoral approach, a 6F guiding catheter was placed into the right innominate artery. ", "Subsequently, a microcatheter with a 0.014‐in microguidewire was navigated into the subclavian artery distal to the aneurysm neck. ", "The BDFD delivery system was advanced over the microguidewire using the exchange method. ", "The wire was then removed, and BDFD was deployed across the aneurysm neck using a push‐pull maneuver. ", "After the deployment, balloon angioplasty with a 2.5‐ to 3.5‐mm compliant balloon was performed within BDFD to achieve complete wall apposition of the device.", "\n\nAn additional BDFD was placed in the abdominal aorta across the origin of the lumbar artery using the same procedure. ", "Subsequently, the guiding catheter and sheath were removed and the femoral artery was ligated and allowed to recover. ", "The lumbar artery was used to assess the preservation capability of the intracranial stent with the small branching artery.[15](#jah34432-bib-0015){ref-type=\"ref\"}, [16](#jah34432-bib-0016){ref-type=\"ref\"}, [17](#jah34432-bib-0017){ref-type=\"ref\"}, [18](#jah34432-bib-0018){ref-type=\"ref\"}\n\nEach animal was randomly allocated to the one of the specific time‐point groups for the follow‐up investigation: 1 month (n=5), 3 months (n=5), 6 months (n=5), and 1 year (n=3) after BDFD placement (Figure [2](#jah34432-fig-0002){ref-type=\"fig\"}). ", "After the procedure, antiplatelet therapy with aspirin (30 mg) was continued throughout the study duration.", "\n\n{#jah34432-fig-0002}\n\nAngiographic Analysis {#jah34432-sec-0014}\n---------------------\n\nAfter the observation period, follow‐up angiography was performed and aneurysm occlusion was assessed using a 3‐point scale (complete occlusion, neck remnant, and dome filling). ", "The patency of the branch arteries covered by the device (including the left common carotid artery, right vertebral artery, and lumbar artery) was assessed using a binary scale (patent/occluded). ", "To assess the downstream embolization, any occlusion or thrombus formation at the downstream arteries of the aneurysm, including the distal portion of the subclavian artery, internal mammary artery, brachial artery, and its small branches, was evaluated.", "\n\nOptical Coherence Tomography {#jah34432-sec-0015}\n----------------------------\n\nOptical coherence tomography (OCT) was performed using a commercially available frequency‐domain OCT system (Lunawave; Terumo, Tokyo, Japan). ", "Quantitative analyses were performed at 1‐mm intervals within the device, which was defined as the region between the metallic radiopaque markers. ", "Neointimal thickness was defined as the distance from the center of each strut to the luminal border. ", "Quantitative area measurements in orthogonal angles were performed. ", "The luminal and scaffold areas were drawn by a manual tracing method for each slice. ", "The neointimal area was estimated as the scaffold area minus the luminal area. ", "Luminal area stenosis was calculated by dividing the luminal area by the scaffold area. ", "The appearance of the strut at each time point was visually assessed. ", "Morphometric measurements were performed using the following equations:$$\\text{Neointimal}\\,\\text{area} = \\text{scaffold}\\,\\text{area} - \\text{luminal}\\,\\text{area},$$ $$\\text{Luminal}\\,\\text{area}\\,\\text{stenosis} = 100 - (\\text{luminal}\\,\\text{area}/\\text{scaffold}\\,\\text{area}) \\times 100.$$\n\nQualitative assessment was also performed at 1‐mm intervals within the device for the presence of any mass protruding into the lumen, which reportedly is the sign of thrombosis formation at the luminal surface.[19](#jah34432-bib-0019){ref-type=\"ref\"} The thrombus within the aneurysm dome was excluded for this analysis. ", "Red thrombus was defined by a protruding mass with high backscattering and attenuation, and white thrombus was defined by homogeneous backscattering and low attenuation, according to previous reports.[19](#jah34432-bib-0019){ref-type=\"ref\"} Thrombus formation was assessed using the following equation:$$\\begin{array}{cl}\n & {\\text{Percentage}\\,\\text{thrombus}\\,\\text{formation}} \\\\\n & {= \\left( {\\text{number}\\,\\text{of}\\,\\text{thrombus}\\,\\text{positive\\ sections}} \\right.} \\\\", "\n & {/\\left. {", "\\text{number}\\,\\text{of}\\,\\text{all}\\,\\text{sections}} \\right) \\times 100.} \\\\", "\n\\end{array}$$\n\nTissue Processing {#jah34432-sec-0016}\n-----------------\n\nAfter follow‐up angiography and OCT, the animals were euthanized with a lethal dose of sodium pentobarbital and flushed with saline and 4% paraformaldehyde in phosphate buffered saline. ", "Both the right subclavian artery and abdominal aorta were separately resected en bloc and placed directly in 4% paraformaldehyde in phosphate buffered saline solution for fixation. ", "After overnight fixation, the samples were embedded in paraffin. ", "Axial sections were prepared from the proximal, mid, and distal portions of the scaffold segment. ", "For the subclavian artery, the midportion was set as the aneurysm site. ", "A normal vessel was used as the control. ", "The samples were cut axially at 5 μm using a low‐speed saw.", "\n\nHistopathological Analysis {#jah34432-sec-0017}\n--------------------------\n\nEach sample was stained with hematoxylin and eosin and elastica--van Gieson stains. ", "The tissues within the aneurysm dome were categorized as unorganized thrombus, organized thrombus, and collagenized connective tissue.[15](#jah34432-bib-0015){ref-type=\"ref\"} The vessel injury score was calculated using the Schwartz method.[20](#jah34432-bib-0020){ref-type=\"ref\"} This score has been used to assess vessel injury with the metallic FD.[15](#jah34432-bib-0015){ref-type=\"ref\"}, [16](#jah34432-bib-0016){ref-type=\"ref\"} Calculations were performed using the following equation:$$\\begin{array}{cl}\n & {\\text{Mean}\\,\\text{injury}\\,\\text{score} = \\text{sum}\\,\\text{of}\\,\\text{injury}\\,\\text{scores}\\,\\text{for}\\,\\text{each}\\,\\text{strutr}} \\\\\n & {\\qquad\\qquad/\\text{number}\\,\\text{of}\\,\\text{struts}.} \\\\", "\n\\end{array}$$\n\nAll samples were immunostained by overnight incubation with primary antibodies against CD31 (1:20; Dako, Carpinteria, CA), CD34 (1:500; GeneTex, Irvine, CA), smooth muscle actin (1:100; Novus, Centennial, CO), CD68 (1:100; Abcam, Cambridge, UK), and Ki‐67 (1:100; Dako) at 4°C. ", "Specific binding was visualized using a secondary antibody: Alexa 488 and 594 (Thermo Fisher Scientific, Waltham, MD). ", "ProLong Gold antifade reagent with 4',6‐diamidino‐2‐phenylindole (Thermo Fisher Scientific) was used as a nuclear counterstain. ", "All images were obtained using a fluorescence microscope (BZ‐X800; Keyence, Osaka, Japan). ", "Whole slide imaging was performed for image analysis. ", "All parts of the specimen were sequentially imaged at ×20 magnification; subsequently, the images were merged. ", "After taking an image of the complete slide, the number of CD68‐ and Ki‐67--positive cells within the neointima was counted. ", "The extent of reendothelialization was defined as the percentage of the luminal circumference covered by CD31‐positive cells after manual tracing. ", "All imaging analyses were performed using a BZ‐X800 analyzer (Keyence). ", "The following equations were used for immunohistochemical analysis:$$\\text{Inflammation}\\,\\text{of}\\,\\text{the}\\,\\text{neointima} = \\text{number}\\,\\text{of}\\,\\text{CD68‐positive}\\,\\text{cells\\ within}\\,\\text{the}\\,\\text{neointima},$$ $$\\text{Neointimal}\\,\\text{cellular}\\,\\text{proliferation} = \\text{number}\\,\\text{of}\\,\\text{Ki67‐positive}\\,\\text{cells}\\,\\text{within}\\,\\text{the}\\,\\text{neointima},$$ $$\\begin{array}{cl}\n & {\\text{Re‐endothelialization} =} \\\\\n & {(\\text{circumference}\\,\\text{of}\\,\\text{the}\\,\\text{CD31‐positive}\\,\\text{portion}\\,\\text{of}\\,\\text{the}\\,\\text{lumen})/} \\\\\n & {(\\text{total}\\,\\text{luminal\\ circumference}) \\times 100.} \\\\", "\n\\end{array}$$\n\nScanning Electron Microscopy {#jah34432-sec-0018}\n----------------------------\n\nAfter follow‐up angiography and OCT, one subject per each time point was used for scanning electron microscopy. ", "The subject was first deeply anesthetized and then pressure perfused with saline, followed by a fixative solution of 4% paraformaldehyde. ", "The samples were then harvested and fixed with 4% paraformaldehyde and 2% glutaraldehyde at 4°C overnight. ", "After fixation, the arteries were incised at the long‐axis direction of the vessel and gently opened to expose the aneurysm neck and the ostium of the lumbar artery. ", "After fixation with 1% OsO~4~ for 2 hours, the specimens were dehydrated, dried, and coated with a thin layer of platinum palladium. ", "The specimens were then examined using a Hitachi S‐4700 scanning electron microscope (Hitachi, Tokyo, Japan). ", "Neointimal formation on the luminal surface, patency of the branching arteries, and occurrence of strut fragmentation were examined.", "\n\nStatistical Analysis {#jah34432-sec-0019}\n--------------------\n\nContinuous variables are expressed as the mean±SD or median with interquartile range, depending on the distribution of the variable. ", "The normality of distribution was assessed using a quantile‐quantile plot and Shapiro‐Wilk test.", "\n\nComparisons of the results of angiography, OCT, and histological examination over time were performed using a linear mixed effect model that evaluates the effect of time. ", "Although the same animal was not evaluated at different time points, multiple short‐axis measurements from the arterial segment were obtained with the OCT and histological analyses. ", "Because a repeated assessment was performed for these analyses, the animal was specified as a random effect to account for the effects that individual animals might have on the results. ", "When the distribution of the results was different from the normal distribution, the logarithm of the data was used. ", "According to the relatively small sample size at each time point, the whole group was divided into 2 groups: the 1‐ to 3‐month group was compared with the 6‐ to 12‐month group. ", "Two‐sided *P* values of \\<0.05 were considered statistically significant and expressed at 2 significance levels. ", "All statistical analyses were performed using R statistical software (version 3.0.2) and SAS (version 9.4).", "\n\nResults {#jah34432-sec-0020}\n=======\n\nAll experimental procedures were successfully performed. ", "After the experiment, one animal allocated to the 6‐month follow‐up group died because of a severe wound infection and was excluded from the analysis.", "\n\nPolymer Degradation {#jah34432-sec-0021}\n-------------------\n\nThe results are summarized in Figure [1](#jah34432-fig-0001){ref-type=\"fig\"}. ", "The initial Mw of the polymer was 185 000 g/mol. ", "On real‐time degradation analysis, we found a decrease in Mw of 15% at 91 days (3 months), 45% at 273 days (9 months), 83% at 364 days (1 year), and 95% at 553 days (1.5 years) (Figure [1](#jah34432-fig-0001){ref-type=\"fig\"}D).", "\n\nOn accelerated degradation analysis, Mw decreased by 81% at 28 days and by 95% at 35 days. ", "According to the differential scanning calorimetry analysis, the melting point decreased in accordance with the change in Mw: 187.5°C at 3 days, 183.4°C at 14 days, and 161.2°C at 42 days. ", "Crystallinity was consistent throughout degradation: 74.2% at 3 days, 75.2% at 14 days, and 74.3% at 42 days (Figure [1](#jah34432-fig-0001){ref-type=\"fig\"}C).", "\n\nAngiographic Outcomes {#jah34432-sec-0022}\n---------------------\n\nThe results are summarized in the [Table](#jah34432-tbl-0001){ref-type=\"table\"} and Figure [3](#jah34432-fig-0003){ref-type=\"fig\"}. ", "The location of the device was determined by the gold markers at both ends. ", "The mean diameter of the parent artery was 2.9±0.5 mm. ", "Aneurysm occlusion rates were 0% (0/5) at 1 month, 20% (1/5) at 3 months, 50% (2/4) at 6 months, and 33% (1/3) at 1 year. ", "All branching arteries (4 left common carotid arteries, 3 vertebral arteries, and 17 lumbar arteries) were found patent at all time points (24/24, 100%). ", "There was no downstream arterial occlusion or thrombus formation at the downstream arteries of the aneurysm.", "\n\n###### \n\nMorphometric Measurements at Various Time Points\n\n Variable 1 mo (n=5) 3 mo (n=5) 6 mo (n=4) 1 y (n=3)\n --------------------------------------- ------------ ------------ ------------ -----------\n Aneurysm morphology, mean±SD, mm \n Height 3.6±1.0 5.6±1.5 5.8±2.1 5.3±1.3\n Width 1.8±0.8 2.5±1.2 2.4±0.5 3.2±1.1\n Neck 2.3±0.6 2.6±1.3 3.6±0.7 3.9±1.6\n Aneurysm occlusion, n/total (%) \n Complete occlusion 0/5 (0) 1/5 (20) 2/4 (50) 1/3 (33)\n Neck remnant 3/5 (60) 2/5 (40) 2/4 (50) 0/3 (0)\n Branching artery patency, n/total (%) \n CCA 1/1 (100) 2/2 (100) 0/0 (N/A) 1/1 (100)\n VA 2/2 (100) 1/1 (100) 0/0 (N/A) 0/0 (N/A)\n Lumbar artery 5/5 (100) 5/5 (100) 4/4 (100) 3/3 (100)\n\nCCA indicates common carotid artery; N/A, not available; VA, vertebral artery.", "\n\n{#jah34432-fig-0003}\n\nOCT Outcomes {#jah34432-sec-0023}\n------------\n\nThe results are summarized in Figure [4](#jah34432-fig-0004){ref-type=\"fig\"}. ", "The location of BDFD was determined by the gold markers at both ends in all cases. ", "Neointimal thickness was 70 μm (IQR, 60--90 μm) at 1 month, 90 μm (IQR, 60--120 μm) at 3 months, 80 μm (IQR, 60--100 μm) at 6 months, and 60 μm (IQR, 40--80 μm) at 1 year after BDFD placement. ", "The luminal area stenosis was 13.4% (IQR, 11.2%--16.3%) at 1 month, 16.1% (IQR, 13.1%--19.0%) at 3 months, 11.8% (IQR, 10.3%--13.5%) at 6 months, and 11.4% (IQR, 10.2%--13.3%) at 1 year. ", "The luminal area stenosis was significantly smaller for the 6‐ to 12‐month group than the 1‐ to 3‐month group (*P*=0.01). ", "Because of the little strut thickness, no visible changes in the strut\\'s appearance attributable to polymer degradation were apparent throughout the study. ", "The percentage thrombus formation was 1.9%, 1.8%, 0.0%, and 0.0% at 1, 3, and 6 months and 1 year, respectively. ", "All detected thrombi were relatively small white thrombi, and there was no thrombus at the ostia of the branching vessel.", "\n\n{#jah34432-fig-0004}\n\nHistopathological Outcomes {#jah34432-sec-0024}\n--------------------------\n\nAmong the occluded aneurysms, 1 was filled with a fresh thrombus, and the other 2 were filled with a fresh thrombus and organized tissue (Figure [3](#jah34432-fig-0003){ref-type=\"fig\"}). ", "The remaining 1 was used for scanning electron microscopy analysis and was not subjected to histological examination. ", "The mean vessel injury scores were 0.09±0.20 at 1 month, 0.05±0.10 at 3 months, 0.09±0.19 at 6 months, and 0.03±0.05 at 1 year after BDFD placement; there was no significant difference between the 1‐ to 3‐month and the 6‐ to 12‐month groups in this respect (*P*=0.93).", "\n\nThe mean number of CD68‐positive cells within the neointima was 31.5 (IQR, 8.25--78.75) at 1 month, 41.0 (IQR, 13.6--72.5) at 3 months, 22.5 (IQR, 14.7--58.2) at 6 months, and 9.5 (IQR, 6.25--20.5) at 1 year. ", "Although the inflammatory response seemed to decrease over time after BDFD placement, there was no statistically significant difference between the 1‐ to 3‐month and the 6‐ to 12‐month groups (*P*=0.24). ", "The number of Ki‐67--positive cells within the neointima was 127.5 (IQR, 65.0--328) at 1 month, 86.5 (IQR, 43.2--273.2) at 3 months, 30.0 (IQR, 12.5--139.5) at 6 months, and 27.0 (IQR, 10--69.7) at 1 year. ", "Cell proliferation decreased from the 1‐ to 3‐month group to the 6‐ to 12‐month group (*P*=0.03). ", "The extent of neointimal reendothelialization was 63.0% (IQR, 43.5%--78.3%) at 1 month, 61.1% (IQR, 55.8%--78.3%) at 3 months, 59.7% (IQR, 45.4%--67.7%) at 6 months, and 53.8% (IQR, 22.9%--69.3%) at 1 year. ", "There was no significant difference in reendothelialization extent between the 1‐ to 3‐month and the 6‐ to 12‐month groups (*P*=0.45). ", "No CD34‐positive cells were observed throughout the experimental period. ", "Immunohistochemistry results are summarized in Figure [5](#jah34432-fig-0005){ref-type=\"fig\"}.", "\n\n{#jah34432-fig-0005}\n\nAll branching arteries covered by BDFD were patent, including the left common carotid, vertebral, and lumbar arteries. ", "No thrombus formed at the branching vessels and at the struts covering the branches. ", "The struts covering these arteries were already surrounded by the neointima tissue at 1 month and made multiple tissue islands. ", "At 1 to 3 months, several macrophages were detected around the struts, and these became rare after 6 months.", "\n\nThe neointima tissue consisted mainly of smooth muscle actin--positive myointimal cells, fibroblasts, and extracellular matrix with elastic and collagen fibers. ", "In addition, the surface of the neointima was covered by the CD31‐positive endothelial cells, although the reendothelialization was not complete. ", "At 1 to 3 months, the CD68‐positive macrophages surrounded the struts and became rare thereafter. ", "There was no eosinophil infiltration or foreign body giant cell within the neointima. ", "Although the myointimal cell layer seemed to not change, the fibroblasts gradually decreased, leaving the collagen fibers.", "\n\nScanning Electron Microscopy Findings {#jah34432-sec-0025}\n-------------------------------------\n\nThe appearance of the internal surface is shown in Figure [6](#jah34432-fig-0006){ref-type=\"fig\"}. ", "At 1 month, most struts were covered by the neointima. ", "In addition, several scattered thrombi were observed on the surface. ", "The thrombi became discreet and gradually reduced over time. ", "We found no apparent strut fracture or fragmentation at the aneurysm ostia and branching vessels.", "\n\n{#jah34432-fig-0006}\n\nDiscussion {#jah34432-sec-0026}\n==========\n\nTo our knowledge, we have reported a device that represents the concept of a BDFD for the first time. ", "Although it is in the early stages of development, the BDFD targets the cerebral aneurysms with anterior circulation, especially the internal carotid artery aneurysms.", "\n\nIn this study, the BDFD almost completely degraded, except for the radiopaque markers at both ends of the device. ", "According to the in vitro degradation test, the BDFD completely degrades at 1.5 years. ", "Because the biodegradable polymer displays similar degradation behavior in vivo and in vitro, it is assumed that the BDFD will completely degrade at 1.5 years in vivo.[21](#jah34432-bib-0021){ref-type=\"ref\"}\n\nWe chose poly‐[l]{.smallcaps}‐lactic acid as the device material because of its biocompatibility, comparatively higher strength and stiffness, and slower absorption rate compared with other polymers.[22](#jah34432-bib-0022){ref-type=\"ref\"} Because of these desirable features, poly‐[l]{.smallcaps}‐lactic acid is the most commonly used platform for biodegradable scaffolds.", "\n\nThe main difference between our device and the currently used coronary biodegradable scaffolds is the strut thickness; although our device had a strut thickness of 40 μm, most coronary biodegradable scaffolds have a strut thickness of \\>100 μm. ", "We could reduce the strut thickness because a much less radial force is required for intracranial aneurysms compared with the coronary biodegradable scaffold.[23](#jah34432-bib-0023){ref-type=\"ref\"} This is important because the current coronary biodegradable scaffolds failed to match the clinical results of standard metallic stents because of the higher thrombosis rate attributable to the greater strut thickness.[2](#jah34432-bib-0002){ref-type=\"ref\"}, [3](#jah34432-bib-0003){ref-type=\"ref\"}\n\nAlthough we avoided potential problems associated with excessive strut thickness, there were several concerns during the polymer‐based FD development. ", "In a previous coronary intervention study, biodegradable polymeric stents induced more prominent neointimal formation and a greater inflammatory response than metallic stents did, which persisted for \\>3 years.[24](#jah34432-bib-0024){ref-type=\"ref\"} The present study focused on investigating the feasibility of this device.", "\n\nThe vessel morphological characteristics were assessed mainly using OCT. ", "Contrary to our apprehensions, morphological results after device placement were similar to those achieved with a standard FD. ", "Kallmes et al[15](#jah34432-bib-0015){ref-type=\"ref\"}, [16](#jah34432-bib-0016){ref-type=\"ref\"} performed animal experiments with first‐ and second‐generation pipeline embolization devices. ", "The reported injury score was 0.0 to 0.3, the neointima thickness was 100 to 200 μm, and the area stenosis was 8% to 20%. ", "Our results are within the range of their results. ", "In addition, similar temporal morphological changes of the neointima were observed in that the neointima formation peaked after 3 months.[15](#jah34432-bib-0015){ref-type=\"ref\"} Although this study was performed under single antiplatelet therapy, thrombus formation at the BDFD surface was rare, and no thrombus was detected after 6 months. ", "Matsuda et al[19](#jah34432-bib-0019){ref-type=\"ref\"} reported that the metallic FD (Pipeline Flex) under single antiplatelet therapy displayed extensive thrombus formation at 7 days after device placement. ", "Conversely, Marosfoi et al[25](#jah34432-bib-0025){ref-type=\"ref\"} reported no significant difference in the degree of acute thrombus formation between single or dual antiplatelet therapy, but there was a significant difference between the standard FD and the FD with phosphorylcholine coating. ", "To the best of our knowledge, there is no report on thrombus formation with metallic FDs at the late time period, and direct comparison is required to assess whether there is a difference between metallic FDs and those with biodegradable polymers.", "\n\nThe biological response of the vessel wall was assessed using histological analyses. ", "On the basis of the immunohistochemical analysis of Ki‐67, although the implantation of the device induced some neointimal proliferation at 1 to 3 months, these responses attenuated after 6 months. ", "This result was compatible with the OCT findings, which showed that lumen area stenosis decreased after 6 months. ", "On the other hand, the number of CD68‐positive cells was not significantly different from 1 to 3 months to 6 to 12 months, although there was a slight tendency to decrease. ", "There are several studies that suggest a mild but persistent inflammatory response around the strut of the biodegradable vascular scaffold during degradation.[12](#jah34432-bib-0012){ref-type=\"ref\"}, [13](#jah34432-bib-0013){ref-type=\"ref\"}, [24](#jah34432-bib-0024){ref-type=\"ref\"} Because we did not perform a direct comparison with a metallic FD, we were unable to estimate the degree of the inflammatory responses of this device as excessive or not. ", "But we were able to surmise that the persistent inflammation of the device did not result in the excessive neointima proliferation. ", "The neointima tissue consisted mainly of myointima cells (smooth muscle cells) and extracellular matrix, covered by endothelial cells. ", "These histological findings until 1 year were consistent with those of previous reports on the biodegradable scaffold in the cardiovascular literature.[26](#jah34432-bib-0026){ref-type=\"ref\"} Kadirvel et al[27](#jah34432-bib-0027){ref-type=\"ref\"} reported the histological findings after metallic FD placement and found that the neointima consisted mainly of smooth muscle actin--positive smooth muscle cells and the CD31‐positive cells covering them. ", "They also reported that the macrophages and monocytes were found at the struts covering the aneurysm neck.", "\n\nApart from the morphological and biological response, the major concern of the BDFD is the poor visibility under fluoroscope: only the 3 radiopaque markers at both ends are visible. ", "Because the entire scaffold must be well opposed to the vessel wall, we performed angioplasty in every case to ensure wall apposition in this study.[28](#jah34432-bib-0028){ref-type=\"ref\"} To improve the radiopacity, the easiest way is to use the currently used biocompatible radiopaque wires, such as platinum‐tungsten wires; in return, this will reduce the merit of the BDFD. ", "Adding radiopaque compounds to the biodegradable polymer is another way, such as the Fantom Bioresorbable scaffold, which combined iodine and tyrosine‐based polymer (REVA Medical, CA). ", "Some sort of device improvement is strongly required in this aspect.", "\n\nFinally, with respect to device performance, we found the occlusion rate at 6 months to be only 50%, which is lower than that previously reported with FDs (81%--99% at 6 months).[29](#jah34432-bib-0029){ref-type=\"ref\"} One reason for the low rate is that the device often displayed a relatively eccentric strut distribution after deployment, especially at the curved segment. ", "The uneven distribution of the struts may have hindered the extension of the neointimal tissue at the aneurysm neck. ", "Indeed, the device tested in this study was a first‐generation prototype, and the scaffold design has since been modified with a designated 48‐wire braider to improve strut distribution. ", "Experiments are currently testing the efficacy of the newer prototype that exhibits a more even strut distribution. ", "Apart from the device design, there is concern for aneurysm recanalization because the occlusion rate decreased by 50% at 6 months and by 33% at 1 year. ", "Although this decline is merely caused by the relatively small sample size of this study, there is the possible mechanism of aneurysm recanalization inherent to BDFD. ", "During the degradation process, BDFD gradually loses its mechanical strength. ", "When the aneurysm is not completely occluded, this mechanical BDFD weakening can cause device deformation and fracture. ", "Although we did not find any evidence of device fracture by OCT and histology, there remained the possibility of minor device deformation, such as minor strut movement or bending caused by the pulsatile blood flow. ", "Because the tight mesh of the strut (low porosity and high pore density with even distribution) is reported to be an important factor for aneurysm occlusion, these possible minor BDFD deformations at the aneurysm neck can cause aneurysm recanalization, even if the device is not fractured.[30](#jah34432-bib-0030){ref-type=\"ref\"}\n\nIn clinical study, although \\>80% of aneurysms treated by FD achieve complete occlusion at 1 year, some aneurysms do occlude between 1 and 3 years.[10](#jah34432-bib-0010){ref-type=\"ref\"} If the aneurysm does not occlude at 1.5 years after BDFD placement, the aneurysm will have a high risk of recanalization and the risk of distal embolization of the thrombus formed within the dome of the aneurysm. ", "These situations should be well investigated with preclinical studies, and the required duration of degradation with the weakening of the mechanical strength in mind must be well investigated.", "\n\nLimitations {#jah34432-sec-0027}\n-----------\n\nThere are several limitations in this study. ", "First, we did not set up a control group of untreated aneurysms or aneurysms treated with a permanent metallic FD. ", "Untreated aneurysms were not assessed because the long‐term patency of the elastase‐induced rabbit aneurysm model has been previously established.[31](#jah34432-bib-0031){ref-type=\"ref\"} We did not include the current standard FD as a control because this device is a first‐generation prototype and our primary focus was to assess device feasibility for the initial stage of development.", "\n\nSecond, although all branching arteries, including the carotid, vertebral, and lumbar arteries, were preserved in this study, concerns remain about the preservation of the intracranial perforating vessels. ", "The median lumbar artery diameter was 248 μm in this study (IQR, 191--506 μm), and it was different from the intracranial perforating artery diameter, which reportedly ranges from 50 to 840 μm.[32](#jah34432-bib-0032){ref-type=\"ref\"} The preservation capability of the small arteries (\\<150 μm) was not tested in this study.", "\n\nThird, the device requires improvements in several aspects. ", "The tendency of eccentric strut distribution has already been resolved in our subsequent prototype, and we intend to assess the device\\'s performance in another in vivo experiment. ", "The radiopacity of the device is another important aspect, as discussed above, and we are investigating combination of biodegradable polymer and radiopaque materials.", "\n\nFourth, there is an undesirable mechanical behavior peculiar to polymeric materials, which was not mentioned in this study. ", "Polymeric materials are frailer and more vulnerable to creep deformation than metals; this may cause a problem when the device is mounted on a narrow delivery system for a prolonged period. ", "The self‐expanding capacity decreases over time.[33](#jah34432-bib-0033){ref-type=\"ref\"}\n\nFifth, the full degradation period of BDFD was assessed only by the in vitro test. ", "Although the degradation process is based mainly on the nonenzymatic hydrolysis in vivo and in vitro, the flow dynamics, mechanical stress from the vessel wall, and change in pH within the microenvironment also influence the degradation process.[34](#jah34432-bib-0034){ref-type=\"ref\"} The true degradation period should be evaluated with an in vivo model.", "\n\nLast, although there was no evidence of device fragmentation and downstream arterial occlusion or thromboembolism, the tissue of the downstream territory was not assessed histologically for the microembolism. ", "Fracture and fragmentation may occur later because of device weakening attributable to biodegradation. ", "There is a possibility of aneurysm recanalization along with device deformation and fracture, as discussed above; in addition, the fragmented element may cause cerebral infarction. ", "These phenomena should be examined in preclinical studies with longer follow‐ups.", "\n\nSummary {#jah34432-sec-0028}\n=======\n\nThis pilot study supports the concept of BDFDs for the treatment of saccular aneurysms. ", "The analysis performed at the 1‐year follow‐up indicates device feasibility, especially of the morphological aspects. ", "On the other hand, the inflammation response peculiar to the biodegradable scaffold should be further investigated. ", "Our results also indicate the need to improve the device\\'s design in several aspects, especially to increase the aneurysm occlusion capacity. ", "The inherent risks associated with using biodegradable scaffolds were not fully elucidated in this study.", "\n\nSources of Funding {#jah34432-sec-0030}\n==================\n\nThis research is supported by The Translational Research Program; Strategic Promotion for Practical Application of Innovative Medical Technology from Japan Agency for Medical Research and Development.", "\n\nDisclosures {#jah34432-sec-0031}\n===========\n\nNone.", "\n\nThe authors would like to thank Kyoto Medical Planning (Kyoto, Japan) for its generous donation of the device and delivery system. ", "The authors also would like to thank Dr T. Manabe of the Shiga General Hospital for the insightful comments and suggestions on the histopathological analysis.", "\n" ]

{ "pile_set_name": "PubMed Central" }

[ 0, 0.0028735632183908046, 0, 0, 0, 0, 0.027906976744186046, 0.0012033694344163659, 0.006535947712418301, 0.023529411764705882, 0.006802721088435374, 0.002564102564102564, 0, 0.004901960784313725, 0, 0, 0, 0.013513513513513514, 0, 0, 0, 0, 0, 0, 0, 0.015384615384615385, 0.005988023952095809, 0.013157894736842105, 0.006944444444444444, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.011461318051575931, 0, 0, 0.010416666666666666, 0.011627906976744186, 0, 0, 0, 0.00819672131147541, 0.004975124378109453, 0.004739336492890996, 0, 0, 0.011152416356877323, 0, 0, 0.011235955056179775, 0.00980392156862745, 0.006329113924050633, 0.008333333333333333, 0, 0.0074211502782931356, 0, 0, 0, 0.0136986301369863, 0.020512820512820513, 0, 0, 0, 0, 0.004464285714285714, 0, 0, 0, 0, 0, 0, 0, 0.0016181229773462784, 0.0041841004184100415, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.006172839506172839, 0.002797202797202797, 0.013605442176870748, 0.008403361344537815, 0.015625, 0, 0, 0.009009009009009009, 0, 0, 0, 0.001519756838905775, 0.004807692307692308, 0, 0, 0, 0, 0.01818181818181818, 0, 0, 0, 0, 0.005494505494505495, 0, 0, 0, 0, 0.009345794392523364, 0, 0, 0, 0, 0.004405286343612335, 0.010752688172043012, 0.005291005291005291, 0, 0, 0, 0, 0, 0, 0, 0.01008533747090768, 0, 0, 0.013513513513513514, 0, 0, 0, 0.004524886877828055, 0, 0.025906735751295335, 0.0213903743315508, 0, 0, 0, 0, 0, 0, 0, 0.012987012987012988, 0, 0, 0, 0, 0.007462686567164179, 0.018957345971563982, 0.004901960784313725, 0.019417475728155338, 0, 0.01932367149758454, 0, 0, 0, 0, 0, 0.02127659574468085, 0, 0, 0, 0, 0.013157894736842105, 0.005714285714285714, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.021739130434782608, 0, 0, 0.004149377593360996, 0.005988023952095809, 0.008620689655172414, 0.022988505747126436, 0.003436426116838488, 0, 0, 0, 0, 0.007874015748031496, 0, 0, 0, 0, 0.004830917874396135, 0.006779661016949152, 0, 0, 0, 0, 0, 0.0022026431718061676, 0, 0, 0.004424778761061947, 0, 0.005434782608695652, 0.0026455026455026454, 0.005405405405405406, 0, 0.0026455026455026454, 0, 0, 0, 0, 0, 0.01282051282051282, 0, 0, 0.00546448087431694, 0, 0, 0, 0.002583979328165375, 0, 0.0030864197530864196, 0, 0, 0, 0, 0, 0.005780346820809248, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.011450381679389313, 0, 0.007518796992481203, 0.012658227848101266, 0 ]

[ "Filter by:\n\nFeatured Outdoor Shelters & Classrooms\n\nSport and Fitness Playground Equipment\n\nPhysical fitness plays a key role in awarding children a healthy lifestyle, which is imperative to their overall development and growth. ", "Incorporating our gym equipment or trim trail items into your play zone will encourage and allow children to be challenged physically, improving their balance, strength and coordination.", "\n\nFilter by:\n\nFeatured Sport and Fitness Equipment\n\nRole Play\n\nRole play is a key part in every child’s development, and our dens, houses and hideaways are the perfect place for kids to let their imaginations run free. ", "Browse through our amazing collection but remember...we can create bespoke equipment just for you!", "\n\nFilter by:\n\nFeatured Role Play Equipment\n\nOutdoor Sensory Playground Equipment\n\nWhat better way to engage in sensory play than the great outdoors? ", "Children can immerse themselves in a tactile world of fun and learning with our range of Outdoor Sensory play equipment. ", "Our range of products will have something for children of all ages and abilities.", "\n\nFilter by:\n\nFeatured Outdoor Sensory Playground Equipment\n\nOutdoor Playground Equipment\n\nSee our range of outdoor playground equipment, including climbing frames, clamber stacks and trim trails. ", "Let the combination of imaginative play and physical health transform your playground into a natural and engaging active zone.", "\n\nFilter by:\n\nFeatured Outdoor Playground Equipment\n\nSurfacing\n\nOur specially designed playground surfaces create the ideal flooring for all kinds of activities. ", "All our surfaces are durable and safe, making them perfect for any playground. ", "Some are even available in a range of colours to add vibrancy and personality to your site.", "\n\nFilter by:\n\nFeatured Surfacing\n\nStorage and Fencing\n\nKeeping your play zone tidy and safeguarding children is imperative for any school, and our range of fencing and storage allows you to do just that. ", "Keep your play equipment ordered and safe with our hand-crafted storage solutions.", "\n\nFilter by:\n\nFeatured Storage and Fencing\n\nCommercial Playground Equipment\n\nAs one of the UK’s leading suppliers to the Education market, Red Monkey offers a substantial and comprehensive service to all commercial businesses. ", "Having worked with a number of commercial businesses, such as hotels, pubs and parks, we have the experience, knowledge and skill to create unique and solutions." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0, 0, 0, 0, 0.008264462809917356, 0, 0, 0, 0, 0, 0, 0.004901960784313725, 0, 0.013215859030837005, 0 ]

[ "Australia have effectively revealed their full pace attack for the first Ashes Test in Brisbane by withdrawing Josh Hazlewood and Jackson Bird from the third round of the Sheffield Shield in addition to Mitchell Starc and Pat Cummins.", "\n\nThe decision to have Hazlewood train alongside the other three at the National Cricket Centre - rather than playing for New South Wales in the Shield match against Queensland at Allan Border Field next week - followed his impressively rhythmic return to first-class cricket against Western Australia at Hurstville Oval.", "\n\nHaving played in Sydney grade cricket the previous weekend, Hazlewood claimed three wickets in each innings, notably dismissing Test middle-order contender Hilton Cartwright for a first-innings duck, defeating the WA captain Mitchell Marsh in each innings, and also getting the wicket of Shaun Marsh twice. ", "These top-order incisions made it possible for Starc to pluck a hat-trick in each innings while cleaning up the WA tail.", "\n\nFollowing the match, Hazlewood was involved in discussions with coaches, selectors and medical staff, leading to the decision to spell him from the next Shield match, effectively deciding he was ahead of schedule after Hurstville. ", "Instead, he will work alongside Starc, Cummins and Bird under the eye of the Australian bowling coach David Saker in Brisbane. ", "Starc and Cummins were always slated to play the first two Shield games before missing the third.", "\n\n\"The decision on Hazlewood not playing in the third round of the Shield was made in consultation with the player, coaching staff, selectors and team performance staff, following a review of all the bowlers after the last round of the Shield,\" a Cricket Australia spokesperson said.", "\n\nBird's inclusion in the quartet, having bowled well for Tasmania despite their defeats in the opening two matches to Western Australia and Queensland, underlines his long-standing position as the selectors' preferred reserve pace bowler in most conditions. ", "He played the most recent of his Test matches at the MCG against Pakistan on Boxing Day last season, and was a squad member in both India and Bangladesh.", "\n\nBeyond Bird, the most recent other pace bowler in Australian squads has been Chadd Sayers, who was 12th man for last year's day-night Test in Adelaide and is a noted exponent of swing and seam with the pink ball." ]

{ "pile_set_name": "OpenWebText2" }

[ 0.017094017094017096, 0.012461059190031152, 0.012944983818770227, 0.016666666666666666, 0.008583690987124463, 0.015748031496062992, 0.010309278350515464, 0.01060070671378092, 0.007722007722007722, 0.006535947712418301, 0.004672897196261682 ]

[ "MSC Camille\n\nMSC Camille is one of the largest container ships in the world.", "\n\nHull and engine\nThe MSC Camille was built by Daewoo Shipbuilding & Marine Engineering Co in yard number 4136. ", "It is a fully cellular container ship. ", "It can hold up to 14,000 Twenty-foot equivalent units, which makes it one of the biggest container ships in the world. ", "The ship is 365.52m in length and has a 51m beam. ", "MSC Camille as a total output of 98,218 hp. ", "It is also equipped with two tunnel thrusters for better maneuverability in and out of ports.", "\n\nIllegal goods\nA January 2010 inspection of MSC Camille found 1.5 tonnes of illegally imported goods from China to Teesport. ", "Some of them were from animal origin (about 200 boxes), noodles, milk, sweets, and other food. ", "All the items were destroyed.", "\n\nVessel collision\nOn June 20, 2010, Torm Marina (110,000 DWT) was struck by the container ship MSC Camille, 48 miles east of Gibraltar. ", "MSC Camille struck the Torm Marina in the bow, cracking the hull from deck to keel. ", "At time of the collision the Torm Marina had no cargo and no oil stored in effected tanks, avoiding a potential disaster in the Strait of Gibraltar. ", " No injuries were reported.", "\n\nReferences\n\nExternal links\nVideo: MSC Camille\n\nCategory:Container ships\nCategory:2009 ships\nCategory:Ships built by Daewoo Shipbuilding & Marine Engineering" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0.013157894736842105, 0.008928571428571428, 0, 0, 0, 0.022727272727272728, 0, 0.007936507936507936, 0, 0, 0.021897810218978103, 0.011904761904761904, 0, 0, 0.02531645569620253 ]

[ "In this post HPN looks back at the action that took place in the first week of the AFL trade period. ", "For a full recap of all movement on an easy to read webpage, check out Draft Guru.", "\n\nNOTE: We do not use the AFL draft points system to rate players or picks. ", "For more information, see the explaination at the bottom of the post.", "\n\nVerdict: The trade made values McCarthy as having negative value considering the other assets involved. ", "If McCarthy can regain his form from the first half of 2015, Fremantle will have a great deal on their hands.", "\n\nVerdict: Brad Hill is a very polarizing player it seems – more people complained with our rating of Hill than any other rating to date. ", "At just 23, Hill looks to be a solid contributor for the best part of the next decade. ", "HPN rates Fremantle as massive winners here (again).", "\n\nVerdict: A nearly perfectly balanced exchange of picks that seemingly gives both clubs what they want.", "\n\nVerdict: Both clubs receive something that they will be able to get more use out of than the other club could. ", "St Kilda will give the talented Jack Steele plenty of game time, and GWS can get extra bidding point value out of the pick due to the 20% point discount in bid-matching.", "\n\nVerdict: Did this one actually happen? ", "Did Sam Mitchell really leave the Hawks? ", "A club legend moves west to begin his eventual transition to the coaching box, and both sides are reasonable about the value of him. ", "Perhaps the most reasonable trade of the period involving a player so far.", "\n\nVerdict: Hawthorn decided they couldn’t go on without a Mitchell in the middle, and Tom is a massive upgrade to their long term prospects. ", "Hawthorn get an absolute bargain here for the talent young player, giving up only their first round pick for a player likely to produce a lot more value than that pick would have (the pick swap doesn’t matter).", "\n\nVerdict: What were Hawthorn thinking here? ", "Hawthorn pay massive overs in a desperate attempt to get a pick in order to facilitate the O’Meara trade. ", "St Kilda downgrade slightly, and get two good chances this year and one very good one next to further strengthen their list.", "\n\nFree Agents\n\nFor free agent movement, HPN has been tweeting out how the formula values each player on the move. ", "If you’ve got any questions about any aspect of the exchange period, hit us up on Twitter at @hurlingpeople.", "\n\nVickery, as the youngest FA, will likely produce the most going forward, especially in a Hawthorn forward line crying out for a tall target. ", "There’s little doubt that Wells has had the best career of the three, but at 31 it’s unsure how much he will produce going forward. ", "Chris Mayne sits in the middle of the two – a solid forward who contributes in ways not often considered valuable by most standard statistics. ", "As all three were granted the same compensation (a round two pick pick after the club’s own), HPN thinks that the AFL got it right for Vickery and Mayne, and perhaps overcompensated North for Wells.", "\n\nWe’ve got a more detailed post on free agents coming up later today if you want to read more.", "\n\nIMPORTANT: HPN rates trades based on the formulas outlined in this post. ", "The HPN Draft Pick Trade Value Chart (TVC) is compiled from the actual, real-world output of each player selected to at that pick. ", "The HPN Player Value Formula predicts the future output of every AFL player, based on their performance in the past 3 years, their age and their “eliteness”. ", "We discuss the theory behind valuations of players in this post. ", "Both formulas use the same ‘currency’ of future games to measure value, so the points are completely comparable/exchangeable.", "\n\nWe do not use the AFL points system as it doesn’t provide a way to explicitly value players.", "\n\nThe chart above shows the primary difference in our valuations – second round picks have a lot more actual worth because of the output of players drafted there.", "\n\nFinally, note that HPN is not saying the valuations provided here are the only valid considerations for making a trade. ", "Quite often, the point differential in our assessments tell us useful things about what clubs are thinking, valuing and prioritizing during the exchange period." ]

{ "pile_set_name": "OpenWebText2" }

[ 0.009900990099009901, 0.012195121951219513, 0, 0, 0.009433962264150943, 0.01834862385321101, 0.021739130434782608, 0.011494252873563218, 0.038461538461538464, 0.009615384615384616, 0.008849557522123894, 0.011834319526627219, 0, 0.04878048780487805, 0, 0, 0.02127659574468085, 0, 0.022222222222222223, 0, 0.008064516129032258, 0.008771929824561403, 0.009259259259259259, 0, 0.007575757575757576, 0.006993006993006993, 0.020202020202020204, 0, 0.013333333333333334, 0, 0, 0, 0, 0, 0, 0.00819672131147541, 0 ]

[ "New Lenovo™ ideacentre™ Stick 300: A PC that fits in your pocket\n\nThe company’s first compute stick, the Lenovo™ ideacentre™ Stick 300. ", "Designed for consumer usage in the home and on-the-go, the new pocket-sized compute stick is powerful enough for entertainment and web browsing, while affordable enough for existing PC owners to justify owning a second or third mobile PC option. ", "The ideacentre™ Stick 300 expands the mobile computing experience and options for savvy shoppers and is suitable for use in the home, dorm or office, while simultaneously offering portability and convenience for travelers.", "\n\nLenovo’s new compute stick delivers powerful computing and big savings in a small package\n\nConnect Lenovo’s first stick PC to any display to transform it into a computer, ideal for media, web browsing, video chatting and content creation\n\nideacentre™ Stick 300 packs a powerful computing punch for a lightweight budget, starting at US $129.", "\n\nProvides the performance promise of Intel® Atom™ Processor Z3735F and the familiarity of Windows, initially shipping with Windows 8.1 but will be eligible for a free upgrade to Windows 10 starting July 29\n\nThumbs Up on Price\n\nFor the wallet friendly starting price of US $129, this plug and play technology can transform almost any HDMI compatible TV or monitor into a fully functioning Windows-based PC. ", "The ideacentre™ Stick 300 does not look like a traditional computer, but it performs like one once a 2.4GHz wireless keyboard and mouse are added.", "\n\nTransforming the Computing Experience\n\nImagine transforming the traditional TV in a vacation rental into a smart multi-media hub, capable of streaming a movie, video chatting with relatives or editing a work document on the fly. ", "Convert the dusty monitor lurking in a spare bedroom into a web-enabled homework station, or transform a coveted man-cave into a home cinema.", "\n\nWorthy of being crowned the “mini-me” of the traditional desktop PC, the pocket-sized ideacentre™ Stick 300 measures only 15 mm thin, boasts Intel® Atom™ Processor Z3735F, 2G of memory and 32G of storage. ", "It will initially ship with Windows 8.1 but will be available for a free upgrade to Windows 10 starting on July 29. ", "Fully enabled with WiFi and Bluetooth 4.0, the ideacentre™ Stick 300 provides the connectivity you need to power up at any time.", "\n\n“We’ve looked at the computing needs of travelers, business people and families, and realized that a truly portable and affordable solution would be a significant benefit to users of all kinds,” said Jun Ouyang, vice president and general manager, Worldwide Desktop and Visuals, Lenovo. “", "Our goal with the ideacentre™ Stick 300 is to give those users a sense of freedom and enhanced mobility, while packing a serious punch in a small device.”", "\n\nPrice and Availability1\n\nThe Lenovo™ ideacentre™ Stick 300 will be available on www.lenovo.com and will hit retail shelves starting in July. ", "Pricing will start at US $129." ]

{ "pile_set_name": "Pile-CC" }

[ 0.014705882352941176, 0, 0, 0, 0.007371007371007371, 0, 0, 0.0070921985815602835, 0.004830917874396135, 0, 0.0078125, 0.010344827586206896, 0, 0.013986013986013986, 0 ]

[ "Liver ischemia/reperfusion induces an increase of microvascular leukocyte flux, but not heterogeneity of leukocyte trafficking.", "\nLeukocytic response plays a major role in the manifestation of hepatic ischemia/reperfusion (I/R) injury. ", "To clarify whether post-ischemic hepatic leukocyte accumulation is based on increased leukocyte flux to the hepatic tissue due to systemic inflammation or chemoattractant activities or whether it represents solely a local tissue response without changing overall leukocyte flux and trafficking characteristics through the microvasculature, we studied acinar and sinusoidal leukocyte flux and distribution in rat livers in vivo both under normal (sham, n = 8) and post-ischemic (60' ischemia/75' reperfusion) conditions (I/R, n = 8), using fluorescence epi-illumination microscopy (rhodamine-6G). ", "Hepatic ischemia/reperfusion significantly (p < 0.05) increased acinar leukocyte flux (58.4 +/- 20.9 cells/min vs 36.4 +/- 12.8 cells/min in sham controls); however, it did not exhibit increased heterogeneity of acinar leukocyte distribution, as indicated by the unchanged coefficient of variance (CV) of 0.36 +/- 0.16 (sham controls: 0.31 +/- 0.14). ", "In parallel, analysis of individual sinusoidal leukocyte flux demonstrated significantly (p < 0.05) higher values (8.9 +/- 3.7 cells/min) after ischemia/reperfusion when compared with sham controls (5.7 +/- 1.9 cells/min), which, however, was not associated with increased heterogeneity of sinusoidal leukocyte trafficking (CV: 0.85 +/- 0.15 vs 0.85 +/- 0.16 in sham controls) and manifestation of preferential pathways. ", "Analysis of blood cell count did not demonstrate an overall increase of total blood leukocyte count; however, an increased (p < 0.01) fraction of polymorphonuclear leukocytes (65.2 +/- 11.2%) and stab cells (9.5 +/- 7.9%) during post-ischemic reperfusion when compared with sham controls (8.8 +/- 3.5% and 0.2 +/- 0.4%) was demonstrated. ", "Thus, the increase of hepatic leukocyte flux after ischemia/reperfusion may be the result of both the manifestation of a systemic inflammatory response and the increase of local chemoattractant activities, such as the production and release of the cytokine-induced neutrophil chemoattractant of the IL-8 family." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0.002849002849002849, 0, 0, 0 ]

[ "Cytotoxic ester derivatives of the mammary tumor inhibiting antiestrogen 2,3-bis(2-fluoro-4-hydroxyphenyl)-2,3-dimethylbutane.", "\nThe synthesis of the bis-acrylate 12, the bis-beta-chloropropionate 13 and the bis-beta-bromopropionate 14 of the \"partial\" antiestrogen 2,3-bis(2-fluoro-4-hydroxyphenyl)-2,3-dimethylbutane (7) is described. ", "In the case of 13 and 14 the introduction of the beta-haloester-functions moderately reduces the estrogen receptor affinity of 7. ", "However, it was quite strongly diminished in 12. ", "The hydrolytic stability under in vitro-receptor-assay conditions decreases in the order 12 greater than 14 greater than 13. ", "Compared with 7 the estrogenic potency of 12-14 is increased to a great extent. ", "The title compounds cause a strong inhibition of the hormone-dependent MXT-M3.2 mouse mammary tumor." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0.007692307692307693, 0, 0, 0, 0.01 ]

[ "THE news that Grant Stewart has signed his first full-time contract with Glasgow Warriors is a timely reminder of the important role the Premiership has played as a safety net for players who are either late developers or didn’t quite make it through Scottish Rugby development pathway at the first time of asking for some other reason.", "\n\nThe hooker started out in rugby with Strathaven, then moved Dalziel, before joining Glasgow Hawks when he graduated to adult rugby back in 2012. ", "He was involved in the old SRU academy structure, played for Scotland at Under-17, Under-18 and Under-19 levels, and was on the call-up list for the Under-20s during their 2015 Junior World Cup, but by the summer of 2016 his career seemed to have stalled.", "\n\nHe had dropped out of the academy and was mainly a 2nd XV player for Hawks, stuck behind Jack Macfarlane in the pecking order. ", "When Macfarlane moved south to take on a contract with Jersey Reds, Cammy Fenton was lined up for a switch from Stirling County, and only when he was redirected to London Scottish by the SRU did Stewart get his chance.", "\n\nStewart was fortunate that Fin Gillies, a former Warriors hooker himself, was now in charge at Hawks – and he took it upon himself to help the then 21-year-old address the one part of his game which was really holding him back.", "\n\n“Fin was a huge influence, I can’t thank him enough for helping me out when I was there [at Hawks],” he says. “", "He took me under his wing with the throwing which was a struggle of mine and gave me the confidence just to play.", "\n\n“He just gave me some tips and I did a lot of practising during pre-season. ", "Every training session we’d do some one-to-one practice and he helped me out a lot.”", "\n\nJuggernaut\n\nStewart started the 2016-17 season with Hawks like a runaway truck – which was apposite because at that time he was earning a crust working for his family’s haulage business – and before long he was back training with the SRU academy.", "\n\n“He’s a complete anomaly,” said mentor Gillies, back in September 2016. “", "He’s an absolute freak in the gym. ", "He can throw you a 20 metre pass off his left or right hand, kicks probably better than anyone else in the squad, but the only skill he struggled with was above his head at the line-out.", "\n\n“It wasn’t that he couldn’t do it, he just needed to be more confident with it. ", "So, I think that’s what we’ve managed to do, and we’ve tailored our line-out to suit him.”", "\n\nStewart was offered, and accepted, a full-time stage three academy deal at the start of the 2017-18 season, and when the Warriors hit a bit of a hooker injury-crisis midway through last season he got his chance, making his debut off the bench for the club away to Leinster in the Champions Cup last January.", "\n\nThe following week he was called into action after only seven minutes of the team’s final European pool match when George Turner succumbed to a knee injury and played a full part in an impressive 28-21 win over Exeter Chiefs. ", "He picked up five more replacement appearances in the PRO14 during the remainder of last season, returned to Hawks as they successfully fought off the threat of relegation, and was then a surprise call-up to Gregor Townsend’s Scotland squad which toured the Americas during the summer.", "\n\nHe has come a long way in a short period of time.", "\n\n“It’s kind of flown in,” acknowledges Stewart. “", "Around three seasons ago I was playing for Hawks 2nd XV, then the next year I was playing for the 1st XV and now …\n\n“It [becoming a pro] was always at the back of my mind, but I thought it maybe wouldn’t happen – so the plan was to just take it as it comes, keep playing at Hawks, try to get into their 1st XV, then the Scotland Club XV, and keep making the next step.”", "\n\nStewart made his first start for the clubs narrow loss away to Munster at the end of October, and has made four more appearances so far this season off the bench, including in last weekend’s Champions Cup victory over Lyon.", "\n\nOpportunity knocks\n\nTurner was stretchered off the park with an ankle injury in that match and a prognosis has not yet been supplied but he wearing a moon boot at training on Tuesday so his chances of playing this weekend are not great, while Fraser Brown is in a race against time to be back from knee surgery in time for the Six Nations, meaning that Stewart along with Kevin Bryce (who is back at the club on a partnership contract with Stirling County, and focussing again on being a hooker after not really adapting to tight-head prop during two frustrating seasons with Edinburgh) are set to get plenty of game time during the next wee while.", "\n\nAnother academy prospect in Robbie Smith of Ayr may also get a chance to test himself at a higher level, if head coach Dave Rennie continues to resist the temptation to bring in a readymade gap filler.", "\n\nNext up for Warriors is, of course, Edinburgh in the first leg of this season’s 1872 Cup series at Murrayfield on Saturday, which could present Stewart with a chance to test himself against Scotland’s current first choice number two in Stuart McInally.", "\n\n“It’s exciting because he’s kind of The Man at the moment, so I just want to show I can compete with him,” says Stewart. “", "It’s a good chance to show what I can do, and hopefully keep sticking in for chances later on in the season as well.", "\n\n“I’m not sure how long George is out for, but hopefully I can just play every game I can. ", "I’ve definitely got to just take my chance and show the coaches what I can do.”", "\n\nWe hope you have enjoyed reading this article\n\nDavid has worked as a freelance rugby journalist since 2004 covering every level of the game in Scotland for publications including The Sunday Times, The Telegraph, The Scotsman/Scotland on Sunday/Evening News, The Herald/Sunday Herald, The Daily Mail/Mail on Sunday and The Sun.", "\n\nHelp keep Scottish rugby in the news\n\nDAMIEN Hoyland is back from a year-long injury lay-off to take his place in Edinburgh’s starting line-up for tomorrow night’s game against Cardiff, while Mark Bennett is poised to make his return off the bench …" ]

{ "pile_set_name": "Pile-CC" }

[ 0.008928571428571428, 0.006802721088435374, 0.00392156862745098, 0.015503875968992248, 0.01834862385321101, 0.013100436681222707, 0, 0, 0, 0, 0.012096774193548387, 0, 0, 0, 0, 0, 0.006472491909385114, 0.008771929824561403, 0.007017543859649123, 0, 0.02, 0.01084010840108401, 0.013333333333333334, 0.009230769230769232, 0.009852216748768473, 0.015748031496062992, 0.008064516129032258, 0, 0.010869565217391304, 0, 0.018292682926829267, 0.00796812749003984 ]

[ "Share on other sites\n\n+Majesticmerc 872\n\nThe Arch Wiki is great documentation, the best I've ever found for a Linux distro. ", "There's a whole section on installing the most common things (the OS itself, drivers, GUI, etc). ", "They have a beginners guide for people new to Linux and advanced Linux related stuff\n\nShare this post\n\nLink to post\n\nShare on other sites\n\nGerowen 1,150\n\nSwitched from Ubuntu to Debian when Unity became the default login for Ubuntu because for some reason even the \"Classic Gnome\" option felt really slow compared to earlier releases, and I didn't want to regress to an earlier version that wouldn't be supported in 6 months, so I made the move to Debian. ", "I love it.", "\n\nShare this post\n\nLink to post\n\nShare on other sites\n\nGrowled 3,880\n\nI voted Ubuntu. ", "I am using Bodhi Linux, which is based on Ubuntu 10.10 and is using Enlightenment. ", "I love Enlightenment. ", "I have grown to dislike Gnome and Unity and KDE. ", "I don't see myself ever using them again.", "\n\n0\n\nShare this post\n\nLink to post\n\nShare on other sites\n\n+simonlang 530\n\n+simonlang 530\n\nubuntu here! ", "i just shot my fedora 16 installation few mins ago. ", "but all i did was trying to install an ati driver for my x1650 card. ", "booted from installation cd aferwards but no option to repair the installation, also i can only install it after having completely deleted the old one.", "\n\nthats such a no go, i had my small troubles with ubuntu too but i NEVER NEVER NEVER had shot the whole system dead by just installing something wrong.", "\n\nShare this post\n\nLink to post\n\nShare on other sites\n\nJames7 659\n\nI was on Ubuntu Gnome 2 for a long time and then went for a few months with Ubuntu Unity. ", "It was okay but rough around the edges. ", "It felt also even more simplified and less configurable than Gnome, and that was not the direction I wanted to go.", "\n\nLink to post\n\nShare on other sites\n\nHell-In-A-Handbasket 277\n\nUbuntu, only other distros iv used is MadDog on previous PS3, before I switched to Ubuntu on it, and Mandrake 3 long time ago ( mandrake turned me away from *nix as a whole wasn't until Ubuntu 8 I think that I started Using it again\n\nShare this post\n\nLink to post\n\nShare on other sites\n\nKarl L. 275\n\nI started with Ubuntu years ago, but I have used Debian primarily for a couple of years now. ", "I have Debian Stable (Squeeze) on my laptop (one of the many reasons I love Debian is that it is one of the few distributions that still supports the PowerPC architecture as a primary release target), and Debian Testing (Wheezy) on my more powerful desktop (AMD64, not PPC). ", "If I could vote for another distro, I would go with CentOS because I use it on my server due to its stability, support, and SELinux security policy.", "\n\nEdit: For me, the best things about Debian are its rolling release repository (testing), stability, insane architecture support, and its repository utilities (Aptitude and Synaptic)." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0.010309278350515464, 0.0021929824561403508, 0, 0, 0.012048192771084338, 0, 0.04081632653061224, 0, 0, 0, 0, 0, 0, 0.006369426751592357, 0, 0, 0.00437636761487965, 0.014545454545454545, 0.006756756756756757, 0.005434782608695652 ]

[ "Q:\n\nHow to expand movieclips on rollover, then shrink to normal on rollout?", "\n\nI have code to grow my movieClips on mouse over, then shrink back down on mouse out. ", "As far as I can see, my code is correct. ", "I think the issue is that when my cursor is near the edge of the movieclip, both mouse over and mouse out events rapidly fire one after the other, causing the grow/shrink effects to produce a stuttering effect.", "\nI've tried delaying the mouse out to give the mouseover a chance to cause the movieclip to grow past the mouse's mouse out detection... but if the user swipes the cursor across the screen, the various movieclips are often left large, as the mouse leaves the movieclip before the mouse out event is reestablished.", "\nAm I going about this wrong? ", "Here is my code:\nimport mx.transitions.", "Tween; \nimport mx.transitions.easing.*; ", "\n\nstop();\n\nfillClipArray();\n\nfunction fillClipArray() \n{\n for (var prop in this) \n {\n if (this[prop] instanceof MovieClip) \n {\n var mc:MovieClip = this[prop];\n mc.onRollOver = function(){clipOver(this)};\n mc.onRollOut = function(){clipOut(this)};\n }\n }\n}\n\nfunction clipOver(clip:MovieClip)\n{ \n var xScaleX:Tween = new Tween(clip, \"_xscale\", Strong.easeOut, clip._xscale, 125, 0.5, true); \n var xScaleY:Tween = new Tween(clip, \"_yscale\", Strong.easeOut, clip._yscale, 125, 0.5, true);\n}\n\nfunction clipOut(clip:MovieClip)\n{ \n var xScaleX:Tween = new Tween(clip, \"_xscale\", Strong.easeOut, clip._xscale, 85, 0.5, true); \n var xScaleY:Tween = new Tween(clip, \"_yscale\", Strong.easeOut, clip._yscale, 85, 0.5, true);\n}\n\nA:\n\nSeparate your button so the graphic assets are not used for the mouse events.", "\n\nAdd a MovieClip with some shape that roughly matches your button's initial size\nSet its alpha to 0\nGive it an instance name like \"hotspot\"\nAttach your mouse events to that instead of the entire button\n\nThe hotspot will not change size, obviating your problem.", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0, 0, 0, 0, 0.025, 0.010285714285714285, 0, 0 ]

[ "5-Aza-7-deazaguanine\n\n5-Aza-7-deazaguanine or 2-aminoimidazo[1,2-a][1,3,5]triazin-4(1H)-one is a base that is an isomer of guanine. ", "It is used as a nucleobase of hachimoji DNA, in which it pairs with 6-Amino-5-nitropyridin-2-one.", "\n\nReferences \n\nCategory:Nitrogen heterocycles\nCategory:Nucleobases" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0, 0, 0.015151515151515152 ]

[ "Entertainment Reviews:\n\nLos Angeles Times - 10/04/2013\n\"The goofy, absurd horror-comedy BAD MILO! ", "is a cult classic in the making.\"", "\n\nNew York Times - 10/04/2013\n\"Mr. Vaughan is more eager to inspire giggles than gag reflexes....Mr. Marino has intuitive comic timing and a slightly neurotic decency...\"\n\nRogerEbert.com - 10/04/20133 stars out of 4 -- \"On paper, BAD MILO! ", "sounds loosey-goosey-crazy-go-nuts. ", "But in practice, it's understated, if only in relation to its bonkers premise.\"", "\n\nProduct Description:\n\nAn overworked executive befriends the pint-size demon that lives in his colon in hopes of preventing it from attacking his loved ones at will in this gruesomely fun creature feature from executive producers Jay and Mark Duplass. ", "Duncan (Ken Marino) is already under pressure from his mother (Mary Kay Place) and wife (Gillian Jacobs) to start a family when his bullying boss starts to make life at the office a living hell. ", "His gastrointestinal issues becoming too painful to ignore, Duncan visits a hypnotherapist (Peter Stormare), who quickly reveals the cause of his growing discomfort -- a diminutive demon that feeds on anxiety has moved into Duncan's bowels. ", "Now every time Duncan gets stressed, the creature embarks on a bloody rampage. ", "The more it comes out to play, however, the more difficult it becomes to control, prompting worries that it may soon attack his well-intending yet overbearing family. ", "In order to prevent such a disaster, Duncan takes his therapist's advice and befriends the demon, even going to far as to give it a name -- Milo. ", "The more time Duncan spends bonding with Milo, however, the more difficult it becomes to determine exactly where the creature will focus its wrath next. ", "Stephen Root co-stars in a film from writer/director Jacob Vaughan." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0, 0.008333333333333333, 0, 0, 0.007905138339920948, 0.020512820512820513, 0.012448132780082987, 0.012658227848101266, 0, 0.0136986301369863, 0.013071895424836602, 0.029850746268656716 ]

[ "Q:\n\nWhat php version does ubuntu 14.04 officially support?", "\n\nSo simple question in the title, but did not find on google or in there. ", "Or if there is - maybe you can give keywords on how to find this information.", "\n\nA:\n\nI guess I found an answer right after asking question and pasting exactly the title in the google of this question :)\nhttps://wiki.ubuntu.com/TrustyTahr/ReleaseNotes\nSo it is 5.5 as I understand.", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0.004975124378109453, 0 ]

[ "Cancer is Britons' number one fear\n\nNew research has revealed that cancer tops the list of Britons' fears, coming above Alzheimer's disease, heart attack and terrorism.", "\n\nThe survey of 4,000 Britons undertaken by Cancer Research UK also showed there was not enough awareness that a change of lifestyle can help to prevent cancer and that cancer treatment can often be effective if the disease is diagnosed early.", "\n\nAccording to the poll, only one in five people surveyed correctly answered that half of all cancer cases could be prevented by lifestyle changes.", "\n\nAnd two thirds of people underestimated the proportion of cancers that can be prevented.", "\n\nThere are just over 276,000 new cases of cancer in the UK every year and over half could be prevented by the appropriate lifestyle changes.", "\n\nCancer Research UK offered recommendations such as giving up smoking, exercise, a healthier diet and proper sun protection as ways to prevent the disease.", "\n\nThe charity also recommended seeing a doctor whenever any unusual signs are detected." ]

{ "pile_set_name": "Pile-CC" }

[ 0.011904761904761904, 0.00411522633744856, 0, 0, 0, 0, 0 ]

[ "2019 Bassetlaw District Council election\n\nThe 2019 Bassetlaw District Council election took place on 2 May 2019 to elect all 48 members of Bassetlaw District Council in England. ", "This was on the same day as other local elections.", "\n\nThe election resulted in the Labour Party retaining its control of the council with an increased majority. ", "The Conservative Party suffered its worst defeat in Bassetlaw since 1973, winning only 5 seats of the 12 seats it was defending. ", "Both Independents and Labour gained seats from the Conservatives. ", "The Liberal Democrats gained a seat from Labour in East Retford West, the first Liberal Democrat elected in Bassetlaw since 2006.", "\n\nBackground\nBassetlaw is a shire district in Nottinghamshire, England. ", "It is predominantly rural, with two towns: Worksop and Retford. ", "The district was formed in 1974 by the Local Government Act 1972. ", "Local Government in Nottinghamshire is organised on a two-tier basis, with local district councils responsible for local services such as housing, local planning and refuse collection. ", "Nottinghamshire County Council is responsible for \"wide-area\" services, including education, social services and public transport. ", "Bassetlaw District Council has been controlled by Labour for most of its existence, except for a brief period from 2006 to 2011 when it was controlled by the Conservatives.", "\n\nBassetlaw District Council held local elections on 2 May 2019 along with other councils in England as part of the 2019 local elections. ", "Bassetlaw is divided into 25 wards for electoral purposes, with each ward electing between one and three councillors. ", "Until 2015, the council was elected by thirds, with district elections being held every year except the year in which elections to Nottinghamshire County Council took place. ", "The council resolved in 2014 to hold all-out elections from 2015 onwards, with all 48 councillors elected in all-out elections every four years. ", "Councillors defending their seats in this election were previously elected in 2015. ", "At that election, Labour won a majority of seats and retained its control of the council, which it has controlled since 2011.", "\n\nResult\nThe election resulted in the Labour Party retaining its control of the council\n\nWard results\nAsterisks denote incumbent Councillors seeking re-election.", "\n\nBeckingham\n\nBlyth\n\nCarlton\n\nClayworth\n\nEast Markham\n\nEast Retford East\n\nEast Retford North\n\nEast Retford South\n\nEast Retford West\n\nEverton\n\nHarworth\n\nLangold\n\nMisterton\n\nRampton\n\nRanskill\n\nSturton\n\nSutton\n\nTuxford and Trent\n\nWelbeck\n\nWorksop East\n\nWorksop North\n\nWorksop North East\n\nWorksop North West\n\nWorksop South\n\nWorksop South East\n\n† Colin Barton was expelled from the Conservative Party in April 2019, however he remained a nominated Conservative candidate in this election as the deadline for withdrawals had passed.", "\n\nReferences\n\nCategory:2019 English local elections\n2019\nCategory:2010s in Nottinghamshire\nCategory:May 2019 events in the United Kingdom" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0.016853932584269662, 0, 0.009174311926605505, 0.007751937984496124, 0.015151515151515152, 0, 0, 0, 0, 0, 0.007633587786259542, 0.005813953488372093, 0.007246376811594203, 0, 0.005747126436781609, 0, 0, 0, 0.006211180124223602, 0.005703422053231939, 0 ]

[ "St. Maurice Plantation\n\nSt. Maurice Plantation is a historic mansion on a plantation off the banks of the Red River of the South in Winn Parish, Louisiana, USA.", "\n\nHistory\nThe house was built for the Prothro family, and it was completed in 1845. ", "It was designed in the Greek Revival architectural style. ", "By 1850, members of the Prothro family and their African slaves died of the yellow fever. ", "A decade later, during the American Civil War of 1861-1865, the plantation was taken over by the Union Army.", "\n\nIt has been listed on the National Register of Historic Places since April 3, 1979. ", "It was destroyed by fire on June 5, 1981. ", "It was removed from the National Register in January 2019.", "\n\nReferences\n\nCategory:Houses on the National Register of Historic Places in Louisiana\nCategory:Greek Revival architecture in Louisiana\nCategory:Houses completed in 1845\nCategory:Buildings and structures in Winn Parish, Louisiana\nCategory:Antebellum architecture\nCategory:Plantation houses in Louisiana" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0.00625, 0.011904761904761904, 0, 0.011111111111111112, 0.018518518518518517, 0.011627906976744186, 0, 0.017241379310344827, 0.006622516556291391 ]

[ "Another man identified as Majid called it “cheap and extremely inappropriate”. ", "A third said it was “disgusting”….", "\n\nUnder Saudi Arabia’s strict interpretation of Sharia, women are barred from obtaining driving licences and those who flout the ban are jailed. ", "All female citizens are subject to a crushing guardianship system that obliges them to seek permission from male relatives to do everything from opening a bank account to travelling.", "\n\nMagic Blue Smoke\n\nHouse Rules:\n\n1.) ", "Carry out your own dead.", "\n2.) ", "No opium smoking in the elevators.", "\n3.) ", "In Competitions, during gunfire or while bombs are falling, players may take cover without penalty for ceasing play.", "\n4.) ", "A player whose stroke is affected by the simultaneous explosion of a bomb may play another ball from the same place.", "\n4a.) ", "Penalty one stroke.", "\n5.) ", "Pilsner should be in Roman type, and begin with a capital.", "\n6.) ", "Keep Calm and Kill It with Fire.", "\n7.) ", "Spammers will be fed to the Crabipede." ]

{ "pile_set_name": "Pile-CC" }

[ 0.012658227848101266, 0, 0, 0, 0.02631578947368421, 0, 0, 0, 0, 0.008620689655172414, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.02631578947368421 ]

[ "Macrophages normally ingest and kill pathogens in a pro-inflammatory manner that leads to the development of effective immunity, but clear apoptotic cells without stimulating an inflammatory or immune response to self. ", "By contrast, recent data suggest that dendritic cells (DCs) release pro-inflammatory cytokines in response to apoptotic cells. ", "The response to apoptotic cells is normally non-inflammatory and non-immunogenic indicating that the abilities of apoptotic cells to counter pro-inflammatory responses may play a critical and epistatic role in self-tolerance. ", "Perhaps consistent with this notion, apoptotic cells are poorly cleared by macrophages from systemic lupus erythematosus (SLE) patients, can activate macrophages in these patients, and contain within their blebs the nuclear antigens that are recognized by SLE antibodies, which may be factors in disease development. ", "The goals of the application are to address the mechanisms by which apoptotic cells modulate the response of macrophages and DCs to apoptotic cells and inflammatory agonists. ", "The hypothesis is that Toll-like receptors (TLRs) play a critical role in these processes. ", "TLRs are recently identified innate immune receptors that transduce signals in response to lipopolysaccharide (LPS) and other microbial components. ", "TLRs may be important in macrophage down-regulation by apoptotic cells because: 1) apoptotic cells down-regulate macrophage responses to LPS, which signal through TLRs; 2) apoptotic cells and TLRs both regulate macrophage cytokine production; 3) CD14 is important for phagocytosis of apoptotic cells and also enhances TLR signaling; and 4) TLRs are recruited to phagosomes. ", "Aim 1 is the development of reagents for the evaluation of murine TLR expression and function. ", "Aims 2 and 3 uses these reagents to evaluate TLR 2, 4, and 5 expression in macrophages and DCs before and after phagocytosis of apoptotic cells and determines whether apoptotic cells signal through TLRs or modulate the expression and signaling by TLRs." ]

{ "pile_set_name": "NIH ExPorter" }

[ 0, 0, 0, 0.006309148264984227, 0, 0, 0.006756756756756757, 0.00267379679144385, 0, 0 ]

[ "\"Tilbakeblikk\" is the name of a joint project between the Norwegian Forest and Landscape Institute and Norsk Folkemuseum. ", "Tilbakeblikk means “retrospect” or “looking back” in Norwegian, describing the project’s use of photographs taken of the same places separated by long periods of time to illustrate landscape changes in Norway. ", "The images below (starting with photo number two) are interactive—click on each image to see the difference the decades can make." ]

{ "pile_set_name": "OpenWebText2" }

[ 0.03278688524590164, 0, 0 ]

[ "---\ntitle: lipu pi o wawa e lukin\ndate: 2018-10-14\nlanguage: toki pona\n---\n\n# lipu pi o wawa e lukin\n\nsina wile pali e ilo suli la sina wile jo lukin wawa e tawa ala pi tenpo ni.", "\nlukin wawa e tawa ala pi tenpo ni li ilo sina kama e pali ijo pi tenpo pini.", "\nnasin ni li pilin sina ala. ", "sina kama pi toki lawa insa ala e pali ijo pi\ntenpo pini.", "\n\ntenpo ni li ni tenpo.", "\n\ntenpo pini li tenpo ni ala.", "\ntenpo ni la tenpo pini li suli ala.", "\n\ntenpo kama li tenpo ni ala.", "\ntenpo ni la tenpo kama li suli ala.", "\n\ntenpo ni li tawa ale.", "\nsina ken tawa ale e tawa ala pi tenpo ni.", "\n\nsina wile jo tawa ala pi tenpo ni la sina wile tawa ni:\n\n- tenpo mute anu sina pilin ni la sijelo sina suli.", "\n- tenpo mute anu sina pilin ni la sijelo sina lili.", "\n\nsina lukin e ijo mute la sina lukin wawa e nena insa.", "\n\nsina ken tawa ijo mute la sina kepeken tawa ala pi tenpo ni e sina.", "\nsina jo e ni la sina jo lukin wawa pona.", "\nsina jo lukin wawa mute en tawa ala pi tenpo ni ale li pali pona e ilo suli.", "\n\n---\n\nEnglish Translation\n\n# Meditation Document\n\nIf you want to create a large machine, you should learn how to focus on the\nstillness of Now. ", "Focusing on the stillness of now is a tool for you to go back\nto things you were working on before. ", "This method will happen without you\nfeeling anything. ", "You will go back to doing what you were doing before without\nthought.", "\n\nNow is the current time.", "\n\nThe past is not Now.", "\nThe past is not important now.", "\n\nThe future is not Now.", "\nThe future is not important now.", "\n\nNow is always changing.", "\nYou can move with the stillness of Now.", "\n\nIf you want to have the stillness of Now, you want to do this:\n\n- After some time or you feel it, breathe in (expand your chest)\n- After some time or you feel it, breathe out (shrink your chest)\n\nIf you find yourself distracted (looking at many things), focus on the inside of\nyour nasal cavity.", "\n\nYou can do many things if you use the stillness of now. ", "Doing this will let you\nhave focus easier. ", "Lots of focus and the stillness of now help you create a large\nmachine easier.", "\n\n---\n\nThis post is written primarily in [toki pona](http://tokipona.net/tp/Default.aspx),\nor the language of good. ", "It is a constructed language that is minimal (only about\n120 words in total) yet it is enough to express just about every practical day to\nday communication need. [", "It's also small enough there's tokenizers for it](https://github.com/Xe/x/blob/master/web/tokiponatokens/toki_pona.go).", "\n\nHave a good day and be well.", "\n" ]

{ "pile_set_name": "Github" }

[ 0.016853932584269662, 0.025974025974025976, 0.034482758620689655, 0.03508771929824561, 0.043478260869565216, 0.034482758620689655, 0.08333333333333333, 0.034482758620689655, 0.08333333333333333, 0.043478260869565216, 0, 0.02727272727272727, 0.019230769230769232, 0.03636363636363636, 0.014492753623188406, 0.04878048780487805, 0.025974025974025976, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.008620689655172414, 0, 0.008403361344537815, 0, 0 ]

[ "Turkish government has detained Sercan Gökdemir, the deputy governor of Diyarbakır province, and his wife beside of dozens of people in Turkey’s various provinces on the weekend as part of the government’s massive post-coup witch hunt campaign targeting the alleged followers of the Gülen movement after the controversial coup attempt on July 15, 2016.", "\n\nSercan Gökdemir, deputy governor of Diyarbakır province.", "\n\nIt was reported that police teams detained Diyarbakır’s deputy governor Sercan Gökdemir and his wife following detention warrants were issued by Diyarbakır Chief Prosecutor’s Office on Sunday over their alleged links to the Gülen movement.", "\n\nAt least 115 people were detained on Sunday in northwestern Turkey following the detention warrants for 127 people including businessmen, housewives and police officers were issued by the Tekirdag Public Prosecutor’s Office over their alleged links to the Gülen movement and their alleged use of a mobile phone messaging app ByLock.", "\n\nAlso on Sunday, 12 people were detained by police in Mersin province after Mersin Chief Prosecutor’s Office has issued detentions warrants for 20 people over their alleged use of ByLock, a mobile phone messaging app. ", "It was reported that there are public servants, private sector workers and housewives among the detainees.", "\n\nPolice also detained 82 people in Manisa province on Saturday over their alleged links to the Gülen movement after Manisa Chief Prosecutor’s Office has issued detention warrants for 250 people over their alleged use of ByLock. ", "It was reported that among the detainees there are public officials who were previously dismissed by government decrees under the rule of emergency declared immediately after the controversial coup attempt on July 15, 2016.", "\n\nTurkish authorities believe using ByLock is a sign of being a Gülen follower as they see the mobile phone application as the top communication tool among the sympathizers of the Gülen movement. ", "Tens of thousands of civil servants, police officers and businessmen have either been dismissed or arrested for using ByLock since the failed coup attempt in 2016.", "\n\nFour people were also arrested by a court on Saturday as part of a Trabzon-based investigation targeting followers of the faith-based Gülen movement. ", "Four out of eight detained people, including businessmen, teachers and nurses who were earlier dismissed from their jobs by government decrees, were arrested by a Trabzon court on charges of being supporters of Gülen movement. ", "Four others were released on judicial probation.", "\n\nTurkey survived a controversial military coup attempt on July 15 that killed 249 people and wounded. ", "Immediately after the putsch, the Justice and Development Party (AKP) government along with Turkey’s autocratic President Recep Tayyip Erdoğan pinned the blame on the Gülen movement.", "\n\nFethullah Gülen, who inspired the movement, strongly denied having any role in the failed coup and called for an international investigation into it, but President Erdoğan — calling the coup attempt “a gift from God” — and the government initiated a widespread purge aimed at cleansing sympathizers of the movement from within state institutions, dehumanizing its popular figures and putting them in custody.", "\n\nTurkey has suspended or dismissed more than 150,000 judges, teachers, police and civil servants since July 15. ", "Turkey’s Justice Ministry announced on July 13 that 50,510 people have been arrested and 169,013 have been the subject of legal proceedings on coup charges since the failed coup." ]

{ "pile_set_name": "Pile-CC" }

[ 0.005681818181818182, 0.017241379310344827, 0.012448132780082987, 0.008982035928143712, 0.0045662100456621, 0, 0.013100436681222707, 0, 0.015306122448979591, 0.006134969325153374, 0.013157894736842105, 0.004405286343612335, 0, 0, 0.02197802197802198, 0.0024390243902439024, 0, 0.0056179775280898875 ]

[ "Q:\n\nAccessing values from prop.table() in R\n\nI have a probability table as follows:\nprop.table(table(h[[10]]))\n\n 1 12 15 16 3 9 \n0.20000000 0.30000000 0.10000000 0.13333333 0.20000000 0.06666667 \n\nI then converted it to matrix using:\nprobData <- data.matrix(prop.table(table(h[[10]])))\n\nNow I need to run the condition such as:\nif (probData[row,column] > 0.2){\n\n print(print name of that column eg 1,12,15,16) \n\n}\n\nprobData[row,column] gives me the probability.", "The problem is I am not able to access the column names (1,12,15,16) from their respective probabilities.", "\nAny help is appreciated. ", "\n\nA:\n\nYou could call the names directly from the prop.table with a vectorized condition. ", " \nHere's an example with the mtcars data\n(p <- prop.table(table(mtcars[1:10,3])))\n#\n# 108 140.8 146.7 160 167.6 225 258 360 \n# 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.2 \n# \nnames(p[p > 0.1])\n# [1] \"160\" \"360\"\n\nOr in one line, you could do\nnames((p <- prop.table(table(mtcars[1:10,3])))[p > 0.1])\n# [1] \"160\" \"360\"\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0, 0.00303951367781155 ]

[ "After Swedish authorities raided the Pirate Bay back in December, one of the Web’s most popular torrent sites went dark. ", "In January, the site came back with a countdown: People would once again be able to share files—including movies, games, and porn—on Feb. 1.", "\n\nIt appears the Pirate Bay has kept its promise, and even delivered a day early. ", "The site is back on thepiratebay.se. ", "Torrent links are live; however others, including “register” and “contact” links, still give you an error page.", "\n\nThe homepage depicts an orange phoenix—representative of the site rising from the ashes.", "\n\nThe Pirate Bay\n\nThe site could quickly reboot its services thanks to cloud infrastructure and independent virtual machines. ", "When authorities come, it’s easy to pull the plug and start again elsewhere.", "\n\nBut even so, the Pirate Bay’s downtime was much longer than anticipated, giving rise to a number of spin-offs.", "\n\nAlthough the site is back online, Torrent Freak reports that turmoil among the staff has caused a rift, and a number of former staff members have left the site. ", "They are planning on starting their own torrent service and don’t believe the one announced on Saturday is the official service.", "\n\nH/T Torrent Freak | Photo via Oliver Bruchez/Flickr (CC BY SA 2.0)" ]

{ "pile_set_name": "OpenWebText2" }

[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.006134969325153374, 0, 0.014705882352941176 ]

[ "Polymerization characteristics of colored compomers cured with different LED units.", "\nIncomplete polymerization of a resin material used for dental restoration affects the properties of the restoration. ", "We evaluated the structural and mechanical properties of three different colored compomers cured with three different LED units to observe the characteristic changes occurring in different matches. ", "Polytetrafluoroethylene molds (5 mm in diameter and 2 mm in thickness) were used to prepare disk-shaped sample specimens. ", "Nine sample groups (each of five specimens) were prepared, three each of different compomer resin colors (gold, berry, and silver). ", "Samples were cured using three different LED units (Optima, VALO, and Demi Ultra), according to the manufacturers' instructions. ", "Microstructural properties of samples were characterized by determining the degree of curing using Fourier-transform infrared spectroscopy and by analyzing sample morphology under a scanning electron microscope. ", "The Vickers hardness, compressive strength, and elastic modulus of the samples were measured to investigate their mechanical properties. ", "The degrees of curing decreased in the order of silver > berry > gold for all curing units. ", "Conversely, gold compomers exhibited poorer mechanical values than the berry and silver counterparts. ", "The Optima 10 unit yielded slightly higher degrees of curing than the other devices, followed by Demi Ultra and VALO light-curing units, respectively. ", "The resin color affected the structural and mechanical properties of the compomers, possibly as a result of the complex interactions and relationships between the irradiation light and resin material, such as light absorbance and reflectance; thus, depending on the color as well as the curing protocol." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0, 0, 0.015503875968992248, 0, 0.0072992700729927005, 0, 0, 0.013245033112582781, 0 ]

[ "German travel & tourism Blog\n\nThe latest news from Germany is Wunderbar!", "\n\nBread, glorious bread!", "\n\nIt is such an intrinsic part of society it deserves to be listed by UNESCO, bakers say\n\nGerman bread is the greatest thing since, well … sliced bread! ", "The country boasts more than 3,000 different varieties of the stuff, far more than any other country in the world. ", "Whether white, grey or black, made with pumpkin or with sunflower seeds, whether bread roll or pretzl – it doesn’t just fill the stomach, it’s a work of art and an intrinsic part of German culture. ", "And despite the onslaught of supermarkets, huge numbers of villages still have their own bakery.", "\n\nThis vast offer is partly down to Germany’s varied climate, which allows wheat, rye, barley and oats but also buckwheat, maize or millet and potatoes to thrive. ", "All are used as the basis for dough.", "\n\nBut history and geography were equally important to diversity, as the small princedoms of the olden days sought to carve out their own identity with individual coins… and bread varieties. ", "In true German fashion, the bakers huddled together in guilds and passed their craft on from generation to generation. ", "Today, most regions still have their own local speciality, such as the Westphalian Pumpernickel, Bavarian Pretzl or Oldenburger Brot, a wholemeal bread made of rye. ", "And German bakers and patissiers are found in big cities and hotels all over the world.", "\n\nIf we really needed proof that Germans know which side their bread (brot) is buttered we’d find it in expressions such as Abendbrot (supper), Brötchengeber (employer) or Brotneid (envy) and also in valued institutions such as the Museum of Bread Culture in Ulm. ", "In the fact that the star puppet of German children’s TV channel KI.KA is ‘Bernd, das Brot’ – a slice of talking bread. ", "Or the “Day of the Bread” on 5th May when the Bread Ambassador (2014: celebrity chef Tim Mälzer) promotes the benefits of bread throughout the country.", "\n\nIn fact, it is of such importance to the nation that German bakers are now calling for it to be recognised by UNESCO as part of the “intangible cultural heritage”, just like the Argentian tango or Chinese acupuncture." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0, 0.006535947712418301, 0, 0, 0, 0, 0, 0, 0, 0.012121212121212121, 0, 0.015151515151515152, 0.008333333333333333, 0.006622516556291391, 0.0045662100456621 ]

[ "To circ or not to circ: indications, risks, and alternatives to circumcision in the pediatric population with phimosis.", "\nSummary Although there continues to be considerable debate over the merits of circumcision, it is clear that preservation of the pediatric foreskin, even in the presence of phimosis, is a viable option. ", "Steroid topical cream is a painless, less-complicated, and more economical alternative to circumcision for treating phimosis. ", "Success rates are quite high, especially when patient selection is appropriate and parents are adequately instructed on application. ", "In those children in whom topical steroid therapy has failed, there remains a variety of foreskin-preserving surgical options for treating phimosis. ", "Compared to circumcision, these less-invasive techniques are associated with lower morbidities and cost. ", "Furthermore, depending on the tissue-preserving technique used, satisfactory cosmesis is also achieved. ", "Thus, those males who were not circumcised at birth now have medical and surgical options, which will decrease the likelihood of requiring circumcision at an older age. ", "As health care providers in the United States see more and more uncircumcised male children, it is important for these children and their parents to understand the natural history of physiologic phimosis. ", "Additionally, it is the responsibility of health care providers to present the management options available for the treatment of the persistent nonretractile foreskin and/or pathologic phimosis. ", "These options are particularly important for those individuals whose religious, cultural, or personal preference is to retain the foreskin." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0.007518796992481203, 0, 0, 0, 0, 0, 0, 0 ]

[ "mod model;\n\nuse async_graphql::{EmptyMutation, EmptySubscription, Schema};\nuse model::Episode;\nuse slab::Slab;\nuse std::collections::HashMap;\n\npub use model::QueryRoot;\npub type StarWarsSchema = Schema<QueryRoot, EmptyMutation, EmptySubscription>;\n\npub struct StarWarsChar {\n id: &'static str,\n name: &'static str,\n friends: Vec<usize>,\n appears_in: Vec<Episode>,\n home_planet: Option<&'static str>,\n primary_function: Option<&'static str>,\n}\n\npub struct StarWars {\n luke: usize,\n artoo: usize,\n chars: Slab<StarWarsChar>,\n human_data: HashMap<&'static str, usize>,\n droid_data: HashMap<&'static str, usize>,\n}\n\nimpl StarWars {\n #[allow(clippy::new_without_default)]\n pub fn new() -> Self {\n let mut chars = Slab::new();\n\n let luke = chars.insert(StarWarsChar {\n id: \"1000\",\n name: \"Luke Skywalker\",\n friends: vec![],", "\n appears_in: vec![],", "\n home_planet: Some(\"Tatooine\"),\n primary_function: None,\n });\n\n let vader = chars.insert(StarWarsChar {\n id: \"1001\",\n name: \"Luke Skywalker\",\n friends: vec![],", "\n appears_in: vec![],", "\n home_planet: Some(\"Tatooine\"),\n primary_function: None,\n });\n\n let han = chars.insert(StarWarsChar {\n id: \"1002\",\n name: \"Han Solo\",\n friends: vec![],", "\n appears_in: vec![Episode::Empire, Episode::NewHope, Episode::Jedi],\n home_planet: None,\n primary_function: None,\n });\n\n let leia = chars.insert(StarWarsChar {\n id: \"1003\",\n name: \"Leia Organa\",\n friends: vec![],", "\n appears_in: vec![Episode::Empire, Episode::NewHope, Episode::Jedi],\n home_planet: Some(\"Alderaa\"),\n primary_function: None,\n });\n\n let tarkin = chars.insert(StarWarsChar {\n id: \"1004\",\n name: \"Wilhuff Tarkin\",\n friends: vec![],", "\n appears_in: vec![Episode::Empire, Episode::NewHope, Episode::Jedi],\n home_planet: None,\n primary_function: None,\n });\n\n let threepio = chars.insert(StarWarsChar {\n id: \"2000\",\n name: \"C-3PO\",\n friends: vec![],", "\n appears_in: vec![Episode::Empire, Episode::NewHope, Episode::Jedi],\n home_planet: None,\n primary_function: Some(\"Protocol\"),\n });\n\n let artoo = chars.insert(StarWarsChar {\n id: \"2001\",\n name: \"R2-D2\",\n friends: vec![],", "\n appears_in: vec![Episode::Empire, Episode::NewHope, Episode::Jedi],\n home_planet: None,\n primary_function: Some(\"Astromech\"),\n });\n\n chars[luke].friends = vec![han, leia, threepio, artoo];\n chars[vader].friends = vec![tarkin];\n chars[han].friends = vec![luke, leia, artoo];\n chars[leia].friends = vec![luke, han, threepio, artoo];\n chars[tarkin].friends = vec![vader];\n chars[threepio].friends = vec![luke, han, leia, artoo];\n chars[artoo].friends = vec![luke, han, leia];\n\n let mut human_data = HashMap::new();\n human_data.insert(\"1000\", luke);\n human_data.insert(\"1001\", vader);\n human_data.insert(\"1002\", han);\n human_data.insert(\"1003\", leia);\n human_data.insert(\"1004\", tarkin);\n\n let mut droid_data = HashMap::new();\n droid_data.insert(\"2000\", threepio);\n droid_data.insert(\"2001\", artoo);\n\n Self {\n luke,\n artoo,\n chars,\n human_data,\n droid_data,\n }\n }\n\n pub fn human(&self, id: &str) -> Option<usize> {\n self.human_data.get(id).cloned()\n }\n\n pub fn droid(&self, id: &str) -> Option<usize> {\n self.droid_data.get(id).cloned()\n }\n\n pub fn humans(&self) -> Vec<usize> {\n self.human_data.values().cloned().collect()\n }\n\n pub fn droids(&self) -> Vec<usize> {\n self.droid_data.values().cloned().collect()\n }\n}\n" ]

{ "pile_set_name": "Github" }

[ 0.014396456256921373, 0, 0.004366812227074236, 0, 0, 0, 0.003236245954692557, 0, 0, 0 ]

[ "All relevant data are within the paper and its Supporting Information files.", "\n\nIntroduction {#sec001}\n============\n\nG protein-coupled receptors (GPCRs) play many biological functions by activating distinct and diverse signal transduction pathways in different tissues. ", "These receptors have a common seven transmembrane, heptahelical topology and a well-studied mechanism of signaling primarily involving activation of heterotrimeric G proteins \\[[@pone.0187306.ref001]\\]. ", "Compared to the number of studies that examine molecular pharmacological properties, relatively fewer studies have focused on characterizing the cell biological and biochemical properties of GPCRs and the majority of such studies have been with genetically engineered receptors modified with epitope tags \\[[@pone.0187306.ref002]--[@pone.0187306.ref004]\\]. ", "The paucity of studies characterizing the biochemical properties of native receptors is largely due to a lack of selective tools to probe specific receptor functions in endogenous tissues. ", "In order to facilitate studies with GPCRs examining the cell biological, biochemical and anatomical properties of native receptors we decided to generate polyclonal antibodies directed against unique epitopes present in the N-terminal region of GPCRs.", "\n\nPrevious studies have suggested that the N-terminal region of GPCRs is suitable for generation of receptor selective antibodies \\[[@pone.0187306.ref005]--[@pone.0187306.ref008]\\]. ", "For example, by targeting the N-termini of mu and delta opioid receptors, antibodies were generated that we found to be useful for the characterization of the biochemical and cell biological properties of these receptors \\[[@pone.0187306.ref006],[@pone.0187306.ref007]\\]. ", "Moreover, we found that antibodies against certain epitopes differentially recognized activated (agonist-treated) receptors \\[[@pone.0187306.ref006],[@pone.0187306.ref007]\\]. ", "We also showed that these antibodies recognize the post-activation state of the receptor and this enabled us to probe the spatio-temporal dynamics of receptor activation in the brain following peripheral drug administration \\[[@pone.0187306.ref007]\\]. ", "Since antibodies to some (but not all) of the epitopes at the N-terminus of the same receptor exhibited this property, we proposed that the N-terminal region undergoes a conformational change in response to activation that leads to the movement of the glycosylated side chains on residues in the N-terminus resulting in revealing or masking of the epitope \\[[@pone.0187306.ref006],[@pone.0187306.ref007]\\]. ", "We tested this by generating antibodies to the N-terminal region of 6 different family A GPCRs, and found that a portion of the receptor N terminus is masked, and another portion is unmasked upon agonist-induced receptor activation \\[[@pone.0187306.ref006]\\]. ", "We also found that the conformation-sensitive antibodies could be used to examine the duration and extent of activation of endogenous receptors \\[[@pone.0187306.ref006]\\]. ", "From this we surmised that targeting a specific region in the N-terminus that is proximal to residues that have the potential to be glycosylated (Asn or Ser/Thr) should result in the generation receptor-selective antibodies that recognize conformational changes in native receptors in endogenous tissue. ", "In this study we present data with 38 antibodies to 34 different family A GPCRs that are able to recognize native receptors. ", "When examined for their ability to differentially recognize activated receptors, we found that the majority (but not all) of the antibodies exhibited increased recognition of agonist-treated receptors; none exhibited decreased recognition. ", "Next we analyzed the 38 epitopes in order to predict the optimal epitope sequence that would be differentially recognized by the antibody. ", "Finally, since the highest change in recognition was observed with D2 dopamine receptors we characterized this antibody further. ", "We find that the antibody to D2 dopamine receptors is able to detect the dose-dependent increases in receptor recognition in heterologous cells, in cell lines expressing endogenous receptors and in primary neurons. ", "Furthermore, we find that peripheral administration of cocaine (a dopamine transporter blocker that causes increases in the local concentration of dopamine at the synapse and hence activation of dopamine receptors) leads to a significant increase in the D2 receptor recognition in select brain regions. ", "Taken together these studies show that conformation-sensitive antibodies would also be helpful in the detection of receptors from endogenous tissues and in examining dynamics of receptor activation following drug administration.", "\n\nMaterials and methods {#sec002}\n=====================\n\nGeneration of antibodies {#sec003}\n------------------------\n\n38 different antigenic peptides (shown in [Table 1](#pone.0187306.t001){ref-type=\"table\"}) were used to generate antisera against 34 different GPCRs. ", "These peptides were commercially synthesized on a poly-lysine backbone as multiple antigenic peptides (MAPs) \\[[@pone.0187306.ref006]\\] by the Tufts University Core Facility. ", "Rabbits were used for the generation of antisera. ", "Prior to immunization, 2 ml of blood was collected from the ear vein of each rabbit for the generation of pre-immune sera. ", "Individual MAPS were dissolved in sterile Tris-buffered saline (TBS) (5 mg/0.5 ml). ", "An emulsion was prepared by placing equal volumes of alumen and MAP solution in an eppendorf tube and vortexing at moderate speed for 20 min. ", "Animals were injected subcutaneously on day 1 with emulsion containing 5 mg MAP, on day 14 with a booster containing 15 mg MAP, on day 28 with a booster containing 25 mg MAP, on day 42 with a booster containing 35 mg MAP. ", "Animals were sacrificed on day 49 by intravenous administration of 0.3 ml pentobarbital through the ear vein. ", "As much blood as possible (\\~60 ml) was collected from the heart and used to prepare serum as described below.", "\n\n10.1371/journal.pone.0187306.t001\n\n###### Receptor recognition by anti-GPCR antibodies following agonist treatment.", "\n\n![](", "pone.0187306.t001){#pone.0187306.t001g}\n\n Receptor Name Antigen AA Agonist \\% of control\n -------- ------------------------------- --------------------- ---- ------------------ ---------------\n **1** **D2 dopamine** `LSWYDDDLER` 10 Quinpirole 153.0 ± 6.0\n **2** **Mu opioid** `SDPLAPASWSPAPGSWL` 18 DAMGO 151.0 ± 7.0\n **3** **Delta opioid** `LVPSARAELQSS` 12 Deltorphin II 150.9 ± 5.0\n **4** **MC4 melanocortin** `NSTHHHGMYTSLHLWN` 16 alpha-MSH 145.9 ± 7.0\n **5** **B1 bradykinin** `QAPANITSCE` 10 BK 1--8 144.3 ± 6.0\n **6** **Beta2 adrenergic** `SRAPDHDVTQE` 11 Isoproterenol 142.0 ± 1.0\n **7** **AT1 angiotensin** `lnsstedgik` 10 Angiotensin II 141.0 ± 3.0\n **8** **GPCR55** `PTLSQLDSN` 9 AM 251 140.5 ± 9.0\n **9** **Ghrelin (GHSR)** `DLDWDASPGN` 10 Ghrelin 140.4 ± 1.0\n **10** **CB1 cannabinoid** `DIQYEDIKGDMA` 11 HU210 139.5 ± 5.0\n **11** **5HT**~**1A**~ **serotonin** `TTTSLEPFGTG` 11 R(+)-8-OH-DPAT 137.3 ± 2.0\n **12** **Neurotensin** `MEATFLALSL` 10 Neurotensin 130.6 ± 6.0\n **13** **Vasopressin V1b** `SEPSWTATPS` 10 Vasopressin 129.7 ± 5.0\n **14** **Cholecystokinin 1** `VVDSLLMNGSNI` 12 CCK8 127.7 ± 2.0\n **15** **Beta1 adrenergic** `PLPDGAATAARL` 12 Isoproterenol 127.6 ± 4.0\n **16** **B2 bradykinin** `NCPDTEWWSWLNA` 13 BK 125.4 ± 2.0\n **17** **A2A adenosine** `GTEAPGGGTRATPYS` 15 UK14304 125.4 ± 6.0\n **18** **M2 muscarinic** `NNGLAITSPY` 10 Bethanecol 125.4 ± 8.0\n **19** **CB2 cannabinoid** `GLEFNPMKEYMI` 11 HU210 125.3 ± 6.0\n **20** **NK1 (substance P)** `VLPMDSDLFP` 10 Substance P 124.5 ± 2.0\n **21** **Neuropeptide Y1** `TLFSRVENYSVHYNV` 15 NPY 124.0 ± 6.0\n **22** **AT2 angiotensin** `DNFSFAATSR` 10 Angiotensin II 119.9 ± 4.0\n **23** **Alpha1b adrenergic** `DLDTGHNTSAPAH` 13 Phenylephrine 116.9 ± 5.0\n **24** **Kappa opioid** `DQQLEPAHISPA` 12 Dynorphin A 116.3 ± 3.0\n **25** **D1 dopamine** `NTSTMDEAGLPAERDF` 16 SKF-38393 113.1 ± 4.0\n **26** **Mu opioid** `SDPLAPASWSPA` 12 DAMGO 101.0 ± 4.0\n **27** **M1 muscarinic** `LAPGKGPWQV` 10 Bethanecol 104.0 ± 6.0\n **28** **CB2 cannabinoid** `CRELELTNGSN` 11 HU210 102.0 ± 4.0\n **29** **Endothelin (EDNRB)** `EVMTPPTKT` 9 ET-1 100.0 ± 8.0\n **30** **NK1 (substance P)** `SNQFVQPTWQ` 10 Substance P 99.0 ± 4.0\n **31** **Leukotriene B4R (BLTR1)** `TAATSSPGGM` 10 Leukotriene B~4~ 100.0 ± 5.0\n **32** **Leukotriene B4R2 (BLTR2)** `SVCYRPPGNE` 10 Leukotriene B~4~ 98.0 ± 7.0\n **33** **Prostaglandin D2** `NESYRCQTSTWV` 12 Prostaglandin D2 103.0 ± 7.0\n **34** **Prostaglandin E2** `NSFNDSRRVE` 10 Prostaglandin E2 103.0 ± 7.0\n **35** **Prostaglandin F2** `SINSSKQPAS` 10 Prostaglandin F2 100.0 ± 4.0\n **36** **Vasopressin V1a** `DVRNEELAKL` 10 Vasopressin 105.0 ± 8.0\n **37** **Vasopressin V2** `STVSAVPG` 8 Vasopressin 101.0 ± 6.0\n **38** **Neurotensin** `LDVNTDIYSKVLVT` 10 Neurotensin 99.0 ± 3.0\n\nMembranes (5 μg) from HEK-293 cells expressing individual receptors were treated with 1 μM of receptor agonist and changes in receptor recognition were ascertained by ELISA using anti-receptor antibodies as described in Methods. ", "Values obtained with vehicle treatment were taken as 100%. ", "Data represent Mean ± SE, n = 3--6.", "\n\nSerum preparation: Blood collected before and after immunization was allowed to clot either for 30--60 min at 37°C or 2 h at room temperature. ", "The clot was separated from the sides of the tube with the help of a Pasteur pipette, kept overnight at 4°C and centrifuged for 10 min at 10,000 x g at 4°C. ", "The supernatant was collected and stored in aliquots at -20°C or at 4°C in the presence of a final concentration of 0.01% sodium azide.", "\n\nAntisera to all 38 MAPs were affinity purified as described below. ", "Whole cell ELISA with HEK293 cells alone or HEK293 cells expressing the various receptors \\[[@pone.0187306.ref009]\\] were carried out to measure receptor recognition by the antibodies.", "\n\nAntibody purification {#sec004}\n---------------------\n\nAll antisera were affinity purified using immobilized epitope peptide column. ", "Briefly, coupling of the immunization peptide to the resin was achieved by incubating 5 mg of the peptide with 4 mL of CarboxyLink Gel slurry (CarboxyLink Coupling Gel (Immobilized Diaminodipropylamine)--Thermo Scientific), 0.5 mL of the Coupling Buffer (0.1M MES buffer \\[2-(N-Morpholino) ethansulfonic acid\\], 0.9% NaCl, pH 4.7) and 60 mg EDC for 3 hours. ", "Affinity purification was carried out as described in CarboxyLink™ Immobilization Kit manufacturer's manual (Thermo Scientific).", "\n\nELISA {#sec005}\n-----\n\nAll affinity-purified antibodies were tested by ELISA assay against the peptide epitope used to generate the antibodies as well as tissue membrane preparations as described previously \\[[@pone.0187306.ref006]\\]. ", "For studies using membrane preparations from cocaine and saline treated animals levels of endogenous D2 dopamine receptors or CB1 cannabinoid receptors were determined by ELISA as described previously \\[[@pone.0187306.ref006]\\].", "\n\nAgonist-mediated change in recognition by the antibody {#sec006}\n------------------------------------------------------\n\nEffect of ligand treatment on receptor recognition by the antibodies was assayed by ELISA using either HEK-293 membranes stably expressing individual receptors or SK-N-SH cells endogenously expressing the receptors being probed. ", "For experiments using membranes, 5 μg/well of membranes were plated onto a 96 well high binding microplate (Greiner Bio-One). ", "The next day, wells were washed with phosphate-buffered saline (PBS) and incubated without or with 1μM of agonist ligands in 50 mM Tris-Cl, pH 7.5, for 30 min at 37°C (in the presence of a protease inhibitor mixture; Sigma). ", "Membranes were quickly rinsed three times (within 5 min) with cold PBS and fixed with 3.7% formaldehyde in 1X PBS for 20 min at room temperature, and then washed 5 times with cold PBS. ", "ELISA was carried out by incubating membranes with 1% BSA and 5% sucrose in PBS for 1 h at 37°C, followed by overnight incubation at 4°C with a 1:500 dilution of primary antibody in 1% BSA in PBS. ", "The wells were then washed three times with PBS (5 min each wash) followed by a 1 h incubation at 37°C with 1:500 dilution (in PBS) of secondary antibody coupled to horseradish peroxidase. ", "The wells were washed three times with PBS (5 min each wash), and color was developed by the addition of the substrate, o-phenylenediamine (5 mg/10 ml in 0.15 M citrate buffer, pH 5, containing 15 μl of H~2~O~2~). ", "Absorbance at 490 nm was measured with a Bio-Rad ELISA reader and values obtained in the absence of ligands were taken as 100%.", "\n\nFor experiments using cells, ELISA was carried out as described previously \\[[@pone.0187306.ref006],[@pone.0187306.ref010]\\]. ", "Briefly, SK-N-SH cells (2 x 10^5^) were seeded into each well of a 24-well polylysine coated plate. ", "Next day cells were treated with different concentrations of receptor agonists in 50 mM Tris-Cl, pH 7.5 for 30 min at 37°C. ", "Cells were quickly rinsed 3 times with ice-cold PBS (to remove the agonist) and fixed with ice-cold methanol for 3 min. ", "ELISA was carried out as described above using 1:1000 dilution of primary antibodies and 1:1000 dilution of secondary antibody coupled to horseradish peroxidase. ", "For experiments examining the effect of antagonist treatment cells were pre-treated with 1 μM antagonist for 10 min at 37°C followed by treatment with 100 nM agonist.", "\n\nWestern blotting {#sec007}\n----------------\n\nWestern blotting analyses were carried out using either whole tissue lysates or membranes preparations as described \\[[@pone.0187306.ref006]\\]. ", "Membranes were prepared from rat tissues (heart, kidney and brain) as described previously \\[[@pone.0187306.ref011]\\]. ", "Primary cortical neurons were generated as described previously \\[[@pone.0187306.ref012]\\] and membranes prepared as described in \\[[@pone.0187306.ref011]\\]. ", "Membranes from a given tissue/cell preparation (10--50 μg) were subjected to SDS-PAGE and transferred to nitrocellulose membranes. ", "Nitrocellulose membranes were cut into strips; individual strips were subjected to Western blot analysis by blocking with 1% albumin in PBS and using a 1:500--1:2000 dilution of a primary antibody (1% albumin in PBS) to a GPCR and a 1:15,000 dilution in PBSt (PBS containing 0.1% Tween20) of IR Dye 700 anti-rabbit IgG (Li-Cor Biosciences, Lincoln, NE). ", "Blots were visualized using the Odyssey Imaging system (Li-Cor Biosciences, Lincoln, NE).", "\n\nAnalysis of the antigenic epitopes {#sec008}\n----------------------------------\n\nThe amino acid representation of unaffected epitopes (epitopes without any antibody recognition or 100% recognition), affected epitopes (epitopes with greater than 100% recognition), the GPCR and exposed residues on each GPCR was calculated by comparing the amino acid frequencies of the epitopes tested to a representative set of proteins from the human proteome. ", "The solvent accessible residues ('exposed residues') on the GPCRs were predicted using the PredictProtein package \\[[@pone.0187306.ref013]\\]. ", "The representative set of human proteins was obtained by using the UniqueProt software \\[[@pone.0187306.ref014]\\] on the human proteome obtained from UniProt. ", "Using the frequencies of each amino acid in the human proteome, we calculated the log base10 fold change by dividing the frequency of a specific amino acid in an epitope to the frequency of a specific amino acid in the representative set and taking the log base 10 of the value \\[[@pone.0187306.ref015]\\] ([Fig 1](#pone.0187306.g001){ref-type=\"fig\"} & [S1 Fig](#pone.0187306.s001){ref-type=\"supplementary-material\"}). ", "To compute the representation of charged residues in the epitopes we tabulated the frequency of negatively, positively, polar and nonpolar residues in the epitopes and compared them to the corresponding frequencies seen in the representative human proteome protein set ([Fig 1B and 1C](#pone.0187306.g001){ref-type=\"fig\"}). ", "We performed a Wilcox-Man-Whitney rank sum test to compare both the distribution of net charges between unaffected epitopes and affected epitopes as well as the lengths of the affected and unaffected epitopes.", "\n\n{#pone.0187306.g001}\n\nCell culture and transfection {#sec009}\n-----------------------------\n\nCHO cells stably expressing Flag-tagged mouse D2 dopamine receptors were grown in F12 media containing 10% FBS, 1% penicillin-streptomycin and 400 μg/ml G418. ", "HEK-293 and SKNSH cells were grown in Dulbecco\\'s modified Eagle\\'s Medium (DMEM) containing 10% FBS and 1% penicillin-streptomycin.", "\n\nAnimals, drug treatment and tissue preparation {#sec010}\n----------------------------------------------\n\n### Mice {#sec011}\n\nAdult male C57Bl/6 mice (Charles River, Inc. 22--24 g at the start of the experiment) were group housed 4--5 per cage with ad libitum access to water and standard mouse chow. ", "Animals were housed under constant temperature and humidity with a 12h light/dark cycle. ", "The animal facility is AAALAC approved and care of animals is provided by a team of veterinarians and trained technicians. ", "Animal use procedures were conducted in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee of Temple University.", "\n\nCocaine hydrochloride, generously provided by the NIDA drug supply program was dissolved in 0.9% saline and administered intraperitoneally (i.p. ", "20 mg/kg body weight). ", "This treatment does not cause suffering or distress to the animals. ", "Mice were sacrificed 30 min later by exposing them to CO~2~ from compressed gas followed by rapid decapitation. ", "Brains were collected, dissected to various regions and membranes prepared as described previously \\[[@pone.0187306.ref016]\\].", "\n\n### Rabbits {#sec012}\n\nMale California rabbits (6 months old) were individually housed with ad libitum access to water and standard rabbit chow. ", "The rabbits were housed under constant temperature and humidity with a 12-hour light/dark cycle. ", "The animal facility is AAALAC approved and care is provided by a team of veterinarians and trained technicians. ", "Rabbits were used to generate polyclonal antibodies to different GPCRs as described above. ", "These procedures do not cause suffering or distress to the animals. ", "Rabbits were sacrificed by intravenous administration of 0.3 ml of pentobarbital through the ear vein and as much blood as possible (\\~60 ml) was collected via cardiac drive.", "\n\n### Rats {#sec013}\n\nMale Wistar rats (6 months old) were group housed 2 per cage with ad libitum access to water and standard rat chow. ", "Animals were housed under constant temperature and humidity with a 12h light/dark cycle. ", "The animal facility is AAALAC approved and care is provided by a team of veterinarians and trained technicians. ", "Rats were anaesthetized (10 mg xylazine/5mg ketamine/100 g i.p) and perfused through the ascending aorta with 50 ml of 0.9% NaCl containing 100 U/ml of heparin, followed by 60 ml of a mixture of 3.75% of acrolein and 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB), pH 7.4 followed by 40 ml of 4% PFA in PB. ", "The brain was then dissected and post-fixed by immersion in 4% PFA in PB for 1 hour at 4°C. ", "Coronal sections (30 μM thick) were cut on a vibratome in ice-cold PB.", "\n\nImmunohistochemistry {#sec014}\n--------------------\n\nFor peroxidase staining: Rat coronal sections in PB were stained as described previously \\[[@pone.0187306.ref017]\\]. ", "Briefly, sections were treated with 1% sodium borohydride in PB for 30 min to block free aldehyde groups. ", "After several washes in PB the sections were incubated in 0.3% H~2~O~2~ in PB to block endogenous peroxidase activity. ", "Sections were washed several times in PB and blocked for 30 min with 6% bovine serum albumin (BSA) and 5% normal goat serum (NGS) in 0.1 M Tris-buffered saline, pH 7.4 (TBS). ", "Sections were incubated overnight at 4°C with 1:1000 or 1:2000 dilution of anti-receptor antibody in TBS containing 3% NGS and 0.3% Triton-X-100. ", "Control sections were incubated with receptor antisera pre-adsorbed with the corresponding immunogenic MAP (1 h prior to use the anti-receptor antibody was incubated with 1 or 5 mg of the corresponding MAP in PBS containing 0.3% Triton-X-100, 3% NGS and a protease inhibitor cocktail). ", "Following incubation with the primary antibody, sections were washed in TBS and incubated for 2 h at RT with biotinylated goat anti-rabbit IgG diluted 1:100 in TBS containing 1% BSA and then for 1 h with streptavidin-biotin peroxidase. ", "After further washes in TBS, sections were incubated with biotinyated tyramine for 10 min and re-incubated in ABC. ", "Sections were then treated with 0.05% diaminobenzidine in the presence of 0.008% nickel chloride and 0.03% H~2~O~2~ in TBS, washed, mounted on gelatin-coated glass slides, dehydrated in graded ethanols, defatted in xylene, cover slipped with Permount and examined using a Leitz Aristoplan microscope.", "\n\nFor fluorescence staining: Rat coronal sections in PB were blocked for 30 min with 6% bovine serum albumin (BSA) and 5% normal goat serum (NGS) in 0.1 M Tris-buffered saline, pH 7.4 (TBS). ", "Sections were incubated overnight at 4°C with 1:1000 or 1:2000 dilution of anti-receptor antibody in TBS containing 3% NGS and 0.3% Triton-X-100. ", "Sections were washed in TBS and incubated for 2 h at RT with 1:1000 dilution of Alexa 488 conjugated secondary antibody in TBS containing 3% NGS and 0.3% Triton-X-100. ", "Sections were washed in TBS, treated with propidium iodide to stain cell nuclei, mounted on glass slides, cover slipped with Permount and examined using a Leitz fluorescence microscope.", "\n\nStatistical analysis {#sec015}\n--------------------\n\nData was analyzed using t-test or One-Way Anova as appropriate using GraphPad Prism 6.0 software. ", "P \\<0.05 was considered to be significant.", "\n\nResults and discussion {#sec016}\n======================\n\nSelection of epitopes and generation of the antibodies {#sec017}\n------------------------------------------------------\n\nSince studies with GPCRs have been severely hampered by a lack of antibodies that selectively recognize native receptors in endogenous tissue, we focused on the N-terminal regions of family A GPCRs and generated antibodies to 34 individual receptors. ", "The criteria applied for the selection of the antigenic epitopes were: (i) location within the N-terminal region, (ii) between 9--17 amino acids in length, (iii) most conserved across species, (iv) relatively hydrophilic in nature, (v) next to (but not containing) glycosylatable residues, (vi) contain or be near one or more proline residues, (vii) not containing cysteine (and if present replaced by serine). ", "The epitopes were chosen with the help of the ExPASy software so that it was possible to predict peptide localization with reference to *N*-glycosylation and phosphorylation sites. ", "The peptides were subjected to a Blast search of the NCBI database to ensure that they represent unique sequences and if possible represent a conserved epitope between species. ", "Antigenic peptides corresponding to residues in rat (conserved in human) receptors were commercially synthesized on a poly-lysine backbone so as to generate multiple antigenic peptides and injected into rabbit. ", "The antibodies were purified using an epitope peptide- affinity column and tested by ELISA and Western blotting (and a few were further characterized by immunohistochemistry). ", "We were able to successfully generate antibodies to all 34 individual receptors although in some cases the antibody titers were substantially lower than expected ([S1 Table](#pone.0187306.s006){ref-type=\"supplementary-material\"}). ", "In a few cases when the first attempt to generate conformational sensitive antibodies failed, we changed the composition of the epitope by either generating a longer epitope (by residue extension) or selecting a different epitope in the N-terminal region; this led to the successful generation of these antibodies ([Table 1](#pone.0187306.t001){ref-type=\"table\"}). ", "Thus conformational sensitive antibodies were obtained following immunization with the `SDPLAPASWSPAPGSWL` peptide but not with the `SDPLAPASWSPA` peptide for mu opioid receptor, `GLEFNPMKEYMI` instead of `CRELELTNGSN` for cannabinoid 2 receptors, `MEATFLALSL` instead of `LDVNTDIYSKVLVT` for neurotensin receptors, and `VLPMDSDLFP` instead of `SNQFVQPTWQ` for neurokinin 1 receptors ([Table 1](#pone.0187306.t001){ref-type=\"table\"}).", "\n\nThe antibodies were purified on an epitope- peptide affinity column and the purified antibodies were characterized as to their ability to recognize the receptor in heterologous cells expressing the receptors; control cells not expressing the receptors were used as negative controls. ", "We find that all of our antibodies recognized the cognate receptors as examined by ELISA assays and only in HEK-293 cells expressing the receptor ([S1 Table](#pone.0187306.s006){ref-type=\"supplementary-material\"}). ", "Next we examined if the antibodies recognized native receptors in endogenous tissues. ", "Western blotting analysis with anti-GPCR antibodies revealed a signal in endogenous tissue ([S2 Fig](#pone.0187306.s002){ref-type=\"supplementary-material\"}); no signal was detected when antibodies were pre-incubated with the immunizing peptide. ", "We further characterized a subset of these antibodies by immunohistochemical analysis and found that these antibodies can detect a signal only in a brain region or tissue expressing the receptors; select examples are shown in [S3](#pone.0187306.s003){ref-type=\"supplementary-material\"} and [S4](#pone.0187306.s004){ref-type=\"supplementary-material\"} Figs. ", "Together these results support the idea that a region within the N-terminal tail of GPCRs is suitable for the generation of receptor specific polyclonal antibodies.", "\n\nAgonist-treatment mediated enhanced recognition {#sec018}\n-----------------------------------------------\n\nIn a previous study we found that the N-terminal region of opioid receptors undergoes a conformational change following agonist treatment, and this leads to increased or decreased receptor recognition by the antibodies \\[[@pone.0187306.ref006],[@pone.0187306.ref007]\\]. ", "In order to see if the antibodies generated in the present study are able to differentially recognize agonist-treated receptors, we subjected membranes from HEK-293 cells expressing individual receptors to agonist treatment and measured the change in recognition of the receptor by the antibody using ELISA. ", "We find that in most cases the antibodies exhibited increased recognition of agonist treated receptors ([Table 1](#pone.0187306.t001){ref-type=\"table\"}). ", "Thus 19 out of the 38 antibodies exhibited \\>25% increase in receptor recognition while 13 antibodies did not exhibit measurable changes in receptor recognition ([Table 1](#pone.0187306.t001){ref-type=\"table\"}). ", "Interestingly, the dopamine D2 receptor antibody showed the highest increase in recognition of more than 50% whereas the dopamine D1 receptor antibody showed only a marginal increase of \\~ 13% ([Table 1](#pone.0187306.t001){ref-type=\"table\"}). ", "Next, we further characterized the set of antibodies with highest increase in receptor recognition using SK-N-SH human neuroblastoma cells endogeneously expressing the receptors. ", "Agonist treatment led to a dose dependent increase in receptor recognition (EC~50~ in the nM range) and this increase in recognition could be blocked by receptor selective antagonists ([Table 2](#pone.0187306.t002){ref-type=\"table\"}, [S5 Fig](#pone.0187306.s005){ref-type=\"supplementary-material\"}). ", "It is possible that these antibodies stabilize the activated conformation of the receptor or function as positive allosteric modulators. ", "It should be mentioned that in our studies incubation with the antibodies was carried out under the conditions of minimal signal transduction (ice-cold buffer); further studies directly evaluating signaling by these antibodies would help clarify whether they function as positive allosteric modulators.", "\n\n10.1371/journal.pone.0187306.t002\n\n###### Endogenous receptor recognition by anti-GPCR antibodies following agonist treatment.", "\n\n![](", "pone.0187306.t002){#pone.0187306.t002g}\n\n --------------------------------------------------------------------------------------------------------------------------------------------------------------------\n Receptor Agonist EC~50~ (nM) E~max~\\ Agonist + antagonist EC~50~ (nM) E~max~ (% control)\n (% control) \n ------------------------------------------------------- ---------------- ------------- ------------- ---------------------------- ------------- --------------------\n D2 dopamine Quinpirole 7.6 ± 1.1 237 ± 8 Quinpirole + Sulpiride n.a. ", " 121 ± 3\n\n 5HT~1A~ serotonin R(+)-8-OH-DPAT 1.4 ± 2.1 221 ± 8 R(+)-8-OH-DPAT + WAY100135 n.a. ", " 116 ± 16\n\n Cholecystokinin 1 CCK8 0.1 ± 0.4 199 ± 17 CCK8 + SR27987 n.a. ", " 129 ± 24\n\n Ghrelin (GHSR) Ghrelin 1.8 ± 1.7 184 ± 5 Ghrelin + YIL n.a. ", " 84 ± 9\n\n Delta opioid[\\*](#t002fn002){ref-type=\"table-fn\"} Deltorphin II 0.6 ± 0.2 182 ± 2 Deltorphin II + TIPPψ n.a. ", " 103 ± 3\n\n Mu opioid DAMGO 1.1 ± 0.3 174 ± 9 DAMGO + CTOP n.a. ", " 98 ± 7\n\n CB1 cannabinoid[\\*](#t002fn002){ref-type=\"table-fn\"} Hu210 4.6 ± 1.3 173 ± 7 Hu210 + SR141716A n.a. ", " 114 ± 8\n\n Beta2 adrenergic[\\*](#t002fn002){ref-type=\"table-fn\"} Isoproterenol 0.9 ± 1.0 172 ± 5 Isoproterenol + CGP12177 n.a. ", " 100 ± 1\n\n B2 bradykinin Bradykinin 5.8 ± 1.9 171 ± 7 Bradykinin + WIN64338 n.a. ", " 89 ± 3\n\n AT1 angiotensin[\\*](#t002fn002){ref-type=\"table-fn\"} Angiotensin II 3.9 ± 1.2 170 ± 10 Angiotensin II + losartan n.a. ", " 108 ± 10\n\n NK1 (substance P) Substance P 0.8 ± 2.0 146 ± 6 Substance P + L733060 n.a. ", " 55 ± 6\n --------------------------------------------------------------------------------------------------------------------------------------------------------------------\n\nSK-N-SH cells were treated with different concentrations (0--10 μM) of receptor agonist in the absence or presence of antagonist and changes in receptor recognition were measured by ELISA using anti-receptor antibodies as described in Methods. ", "Values obtained with vehicle treatment were taken as 100%. ", "Data represent Mean ± SE, n = 3.", "\n\n\\*Values from data in \\[[@pone.0187306.ref006]\\]; n.a., ", "not applicable.", "\n\nAnalysis of the antigenic epitopes {#sec019}\n----------------------------------\n\nNext, we examined whether changes in recognition by the antibody correlated with the structural properties of epitopes such as residue type, overall net-charge as well as length of the epitope ([Fig 1](#pone.0187306.g001){ref-type=\"fig\"} and [S1 Fig](#pone.0187306.s001){ref-type=\"supplementary-material\"}). ", "For this we calculated the amino acid representation of unaffected epitopes (epitopes without increased antibody recognition or 100% recognition), affected epitopes (epitopes with greater than 100% recognition), and exposed residues on each GPCR by comparing the amino acid frequencies of the epitopes tested to a representative set of proteins from the human proteome. ", "We found that, in general, residues within the epitope had a distinctive high representation of Alanine and Tryptophan compared to the residues in a representative set of proteins in the human proteome. ", "They also had a distinctive under representation of positively charged residues (such Lysine and Arginine) compared to solvent-exposed residues on the GPCR. ", "Those epitopes that showed increased recognition were further distinguished by the striking over representation of aspartic acid and by a minor overrepresentation of tyrosine, methionine and histidine residues. ", "Additionally, we detected underrepresentation of glutamine and valine in these epitopes as compared to unaffected epitopes. ", "Plotting the representation of negative, positive, polar and nonpolar (at physiological pH) residues against a representative set of proteins from the human proteome reiterated that affected epitopes (i.e. with increased recognition) contain more negatively charged residues than unaffected epitopes ([Fig 1B](#pone.0187306.g001){ref-type=\"fig\"}). ", "Due to the high abundance of negatively charged residues in the epitopes for antibodies with increased recognition, we plotted the distribution of overall net charge of both affected and unaffected epitopes ([Fig 1C](#pone.0187306.g001){ref-type=\"fig\"}). ", "The epitopes with increased recognition carried a significantly more negative overall net charge compared to the unaffected epitopes (p-value \\<0.001). ", "Finally, the average length of epitopes with increased recognition was seen to be longer, on average, than unaffected epitopes by two residues (p-value = 0.011). ", "Taken together, we find that affected epitopes (i.e. agonist-mediated increased recognition of a GPCR) are on average, twelve residues in length, have an overall net negative charge, and are enriched in residues such as aspartic acid and glutamic acid. ", "These features could potentially be used to identify epitopes in the N-terminal region of other GPCRs in order to generate antibodies that are differentially sensitive to agonist-activated conformational states of receptors.", "\n\nUse of antibodies to detect activity states of endogenous receptors {#sec020}\n-------------------------------------------------------------------\n\nGiven that the antibodies described in this study recognize endogenous receptors we next examined whether they could detect receptor conformational changes in endogenous systems. ", "For this we focused on the dopamine receptors and using SKNSH cells endogenously expressing these receptors or primary cortical neurons we examined the effect of ligand treatment on receptor recognition by the antibodies. ", "We find that treatment with the dopamine D2 receptor (D2R) agonist, quinpirole, but not the antagonist, sulpiride, caused a dose-dependent increase in recognition by the anti-D2R antibodies ([Fig 2B & 2C](#pone.0187306.g002){ref-type=\"fig\"}). ", "Similarly, treatment with the dopamine D1 receptor (D1R) agonist, SKF83822, but not the antagonist, SHC23390, led to an increase in receptor recognition by anti-D1R antibodies ([Fig 2D & 2E](#pone.0187306.g002){ref-type=\"fig\"}). ", "Moreover, we find that antagonist treatment blocked the agonist-mediated increase of receptor recognition ([Fig 2](#pone.0187306.g002){ref-type=\"fig\"}). ", "These results indicate that our antibodies can detect changes in native receptor conformation following ligand treatment.", "\n\n{#pone.0187306.g002}\n\nNext we examined if the anti-D2R antibodies could be used to explore the dynamics of receptor activation *in vivo* following administration of drugs of abuse. ", "For this we examined D2 receptor activation in the brain 30 min after peripheral cocaine administration. ", "Cocaine functions by blocking dopamine transporters thereby enhancing synaptic dopamine levels which then leads to activation of dopamine receptors \\[[@pone.0187306.ref018]\\]. ", "We find that 30 min following cocaine administration, there is a significant increase in antibody recognition of the D2 receptors in the striatum and prefrontal cortex and the extent of this increase is much greater than that observed for CB1R ([Fig 2](#pone.0187306.g002){ref-type=\"fig\"}). ", "These results showing that the D2 receptor activation in the brain can be measured 30 min following peripheral cocaine administration are consistent with data from previous studies that have suggested that the conformation-sensitive antibodies of GPCRs can be used to monitor spatio-temporal dynamics of receptor activity targeting drugs of abuse \\[[@pone.0187306.ref006],[@pone.0187306.ref007]\\].", "\n\nTaken together, our studies show that the criteria that we used for the identification of antigenic epitopes at the N-terminus of GPCRs can be successfully used to generate antibodies that differentially recognize agonist treated receptors. ", "Moreover, these antibodies can be used to detect the presence of activated receptors *in vivo* following peripheral agonist administration. ", "Thus these antibodies could potentially be developed as biosensors for the activity state of a receptor not only in specific brain regions following drug administration but also in different intracellular compartments. ", "The latter type of studies is supported by reports that used nanobodies to show that treatment with an agonist led to the activation of β2-adrenergic receptors not only at the plasma membrane but also in the endosomal compartment \\[[@pone.0187306.ref019]\\].", "\n\nSupporting information {#sec021}\n======================\n\n###### Biophysical features of antigenic epitopes.", "\n\nRepresentation of amino acids in affected epitopes (epitopes with an absorbance greater than 100%), not affected epitopes (epitopes with an absorbance equal to 100%), and residues predicted to be solvent-exposed on the GPCRs tested.", "\n\n(EPS)\n\n###### \n\nClick here for additional data file.", "\n\n###### Western blot analysis showing that anti-GPCR antibodies recognize native receptors.", "\n\nWestern blots with **(A)** rat brain membranes and anti- adenosine A2A (ADORA 2; epitope: `GTEAPGGGTRATPYS`) receptor antibodies; **(B)** rat kidney or brain PVN membranes and either anti-angiotensin 1 (AGTR1; epitope: `lnsstedgik`) or anti-angiotensin 2 (AGTR2; epitope: `DNFSFAATSR`) receptor antibodies; **(C)** rat brain or heart membranes and either anti-α1B (ADRA1B; epitope: `DLDTGHNTSAPAH`), anti-β1 (ADRB1;epitope: `PLPDGAATAARL`) or anti-β2 (ADRB2; epitope: `SRAPDHDVTQE`) adrenergic receptor antibodies; **(D)** rat heart or brain membranes and either anti-bradykinin 1 (BKR1; epitope: `QAPANITSCE`) or anti-bradykinin 2 (BKR2; epitope: `NCPDTEWWSWLNA`) receptor antibodies; **(E)** rat brain membranes and either anti-cannabinoid 1 (CB1R; epitope: `DIQYEDIKGDMA`), anti-cannabinoid 2 (CBR2; epitope: `GLEFNPMKEYM`) receptor or anti-GPR55 (epitope: `PTLSQLDSN`) antibodies; **(F)** rat brain membranes and anti-cholecystokinin (CCK1R; epitope: `VVDSLLMNGSNI`) receptor antibodies; **(G)** rat brain or heart membranes and either anti-dopamine D1 (DRD1; epitope: `NTSTMDEAGLPAERDF`) or anti-dopamine D2 (DRD2; epitope: `LSWYDDDLER`) receptor antibodies; **(H)** rat heart membranes and anti-endothelin B (ET~B~; epitope: `EVMTPPTKT`) receptor antibodies; **(I)** rat heart membranes and anti-ghrelin (GHSR; epitope: `DLDWDASPGN`) receptor antibodies; **(J)** rat heart membranes and either anti-leukotriene B4 receptor 1 (BLTR1; epitope: `TAATSSPGGM`) or anti-leukotriene B4 receptor 2 (BLTR2; epitope: `SVCYRPPGNE`) antibodies; **(K)** rat brain membranes and anti-melanocortin 4 (MC4R; epitope: `NSTHHHGMYTSLHLWN`) receptor antibodies; **(L)** rat brain or heart membranes and either anti-muscarinic M1 (M1R; epitope: `LAPGKGPWQV`) or anti-muscarinic M2 (M2R; epitope: NNGLAITSPY) receptor antibodies; **(M)** rat brain membranes and anti-neuropeptide Y 1 (NPY1R; epitope: `TLFSRVENYSVHYNV`) receptor antibodies; **(N)** rat kidney membranes and anti-neurotensin (NTSR; epitope: `MEATFLALSL`) receptor antibodies; **(O)** rat brain membranes and anti-neurokinin 1 (NKR1; epitope: `VLPMDSDLFP`) receptor antibodies; **(P)** rat brain membranes and either anti-δ opioid (OPRD; epitope: LVPSARAELQSS), anti-κ opioid (OPRK; epitope: DQQLEPAHISPA) or anti-μ opioid (OPRM; epitope `SDPLAPASWSPAPGSWL`) receptor antibodies; **(Q)** rat brain or kidney membranes and either anti-prostaglandin D2 (PTGDR; epitope: `NESYRCQTSTWV`), anti-prostaglandin E2 (PTGER; epitope: `NSFNDSRRVE`) or anti-prostaglandin F2 (PTGFR; epitope `SINSSKQPAS`) receptor antibodies; **(R)** rat brain membranes and anti-serotonin 5HT1A (5HTR1A; epitope: `TTTSLEPFGTG`) receptor antibodies; **(S)** rat kidney or heart membranes and either anti-vasopressin V1a (AVPR1A; epitope: `DVRNEELAKL`), anti-vasopressin V1b (AVPR1B; epitope: `SEPSWTATPS`) or anti-vasopressin V2 (AVPR2; epitope: `STVSAVPG`) receptor antibodies. ", "Representative blots of 3 experiments shown. ", "Lane 1: molecular weight markers; Lane 2; 20--50 μg membrane protein.", "\n\n(EPS)\n\n###### \n\nClick here for additional data file.", "\n\n###### Immunohistochemical evaluation of anti-GPCR antibodies without or with pre-adsorption with corresponding MAP.", "\n\nCoronal sections from rat cerebral cortex (**A**-**D**), rat cerebellum (**E, F**) or rat periaqueductal gray (**G, H**) were incubated with either anti-DRD2 (A, B), anti-CB1R (C, D), anti-OPRD (E, F) or anti-OPRM (**G, H**) antibodies with or without pre-adsorption with the corresponding MAP (5mg/ ml for B, D, F, and 1 mg/ml for H) followed by incubation with HRP-labeled secondary antibodies as described in Methods.", "\n\n(EPS)\n\n###### \n\nClick here for additional data file.", "\n\n###### Immunohistochemical evaluation of anti-GPCR antibodies.", "\n\nCoronal sections from rat hippocampus (**A**) rat nucleus tractus solitarius (**B**), or rat hippocampus (**C**) were incubated with either anti-5HT1AR (**A**), anti-AT1R (**B**), anti-OPRM (**C**) antibodies followed by incubation with Alexa-labeled secondary antibodies (green) as described in Methods. ", "Propidium iodine (red) was used as a nuclear stain. ", "White arrows--red staining associated with receptor staining (green); white arrowheads--red staining not associated with receptor staining.", "\n\n(EPS)\n\n###### \n\nClick here for additional data file.", "\n\n###### Endogenous receptor recognition by anti-GPCR antibodies.", "\n\nSK-N-SH cells were treated with different concentrations (0--10 μM) of either quinpirole (**A**), DPAT (**B**), cholecystokinin 8 (CCK8) (**C**), ghrelin (**D**), DAMGO (**E**), bradykinin (**F**) or substance P (**G**) and changes in receptor recognition were measured by ELISA using either anti-DRD2 (**A**), anti-HTR1A (**B**), anti-CCK1R (**C**), anti-GHSR (**D**), anti-OPRM (**E**), anti-BKR2 (**F**) or anti-NKR1 (**G**) antibodies as described in Methods. ", "Values obtained with vehicle treatment were taken as 100%. ", "Data represent Mean ± SE, n = 3.", "\n\n(EPS)\n\n###### \n\nClick here for additional data file.", "\n\n###### Receptor recognition by anti-GPCR antibodies.", "\n\nHEK293 cells alone (HEK293) or expressing individual receptors (HEK293+ receptor) were subjected to ELISA using anti-receptor antibodies as described in Methods. ", "Data represent Mean ± SE, n = 3--6.", "\n\n(DOCX)\n\n###### \n\nClick here for additional data file.", "\n\n**Financial Disclosure:** This work was supported by National Institutes of Health awards DA008863 and NS026880 to L.A.D., DA013429 and DA018326 to E.M.U., GM108911 to A.S.; R.R. was supported by T32 GM062754.", "\n\n**Competing Interests:** The authors declare that they have no competing interests.", "\n\n[^1]: **Competing Interests:**The authors have declared that no competing interests exist.", "\n" ]

{ "pile_set_name": "PubMed Central" }

[ 0.013157894736842105, 0, 0.0049261083743842365, 0.0056022408963585435, 0, 0, 0.01098901098901099, 0.007352941176470588, 0.011428571428571429, 0.003968253968253968, 0.004914004914004914, 0.007692307692307693, 0.005813953488372093, 0.006578947368421052, 0, 0, 0, 0, 0, 0, 0, 0, 0.011428571428571429, 0, 0, 0, 0.007042253521126761, 0.018018018018018018, 0, 0, 0, 0, 0.005331478905887807, 0, 0.02857142857142857, 0, 0.006369426751592357, 0.007407407407407408, 0, 0.010869565217391304, 0, 0.002793296089385475, 0.015625, 0.008438818565400843, 0.008771929824561403, 0.002840909090909091, 0.007936507936507936, 0.0044444444444444444, 0.021621621621621623, 0.02030456852791878, 0.010582010582010581, 0.009345794392523364, 0, 0.0234375, 0, 0.008064516129032258, 0.008333333333333333, 0.006172839506172839, 0, 0.005235602094240838, 0.01680672268907563, 0.012658227848101266, 0.007633587786259542, 0.01977401129943503, 0.02247191011235955, 0.002232142857142857, 0.014084507042253521, 0.018867924528301886, 0.0023923444976076554, 0, 0.004784688995215311, 0, 0.002577319587628866, 0, 0.007042253521126761, 0.007067137809187279, 0.015151515151515152, 0.0033112582781456954, 0, 0, 0.00975609756097561, 0.006802721088435374, 0, 0, 0.008928571428571428, 0.007936507936507936, 0, 0, 0, 0, 0, 0, 0.007246376811594203, 0, 0, 0.012618296529968454, 0.010869565217391304, 0.014285714285714285, 0.005813953488372093, 0, 0.008403361344537815, 0.005714285714285714, 0.0136986301369863, 0.017482517482517484, 0.00423728813559322, 0.008695652173913044, 0.01, 0.005235602094240838, 0.0136986301369863, 0.011904761904761904, 0.010810810810810811, 0.006535947712418301, 0, 0, 0, 0.0055248618784530384, 0.005649717514124294, 0, 0.005681818181818182, 0.004329004329004329, 0, 0, 0, 0.009302325581395349, 0, 0, 0, 0, 0.005277044854881266, 0.003246753246753247, 0, 0, 0, 0, 0.0033333333333333335, 0, 0, 0, 0, 0.0036407766990291263, 0, 0, 0, 0.006493506493506494, 0, 0, 0, 0, 0.006493506493506494, 0, 0.002336448598130841, 0, 0.03125, 0.017241379310344827, 0, 0, 0, 0.009852216748768473, 0.006369426751592357, 0, 0, 0, 0, 0, 0.006172839506172839, 0, 0, 0, 0, 0, 0.021834061135371178, 0.006535947712418301, 0, 0, 0.014563106796116505, 0.027777777777777776, 0.022727272727272728, 0.004651162790697674, 0.009478672985781991, 0.004784688995215311, 0, 0, 0.003816793893129771, 0, 0.005681818181818182, 0, 0.005037783375314861, 0, 0, 0, 0.0038910505836575876, 0, 0, 0.018518518518518517, 0, 0.009996552912788694, 0, 0, 0.018518518518518517, 0.00847457627118644, 0.002369668246445498, 0.018518518518518517, 0, 0.003257328990228013, 0, 0, 0.018518518518518517, 0, 0.006437768240343348, 0, 0.03125, 0.018518518518518517, 0, 0.006097560975609756, 0.02857142857142857, 0, 0.018957345971563982, 0, 0, 0 ]

[ "<!", "DOCTYPE html>\n<html>\n <head>\n <script src=\"386310-script.js\"></script>\n </head>\n <body>\n <div id=\"test\" style=\"-moz-binding: url(386310-binding.xml#prepend);\">first second third [fourth]</div>\n </body>\n</html>\n" ]

{ "pile_set_name": "Github" }

[ 0, 0 ]

[ "The utility of intraoperative blood flow measurement during aneurysm surgery using an ultrasonic perivascular flow probe.", "\nInadvertent vessel compromise is one major cause of unfavorable outcome from aneurysm surgery. ", "Existing strategies for intraoperative assessment of this complication have potential limitations and disadvantages. ", "We assessed the utility of quantitative intraoperative flow measurements using the Transonic ultrasonic flow probe (Transonic Systems, Inc., Ithaca, NY) during aneurysm surgery. ", "Of all aneurysms treated surgically at our institution from 1998 to 2003, 103 patients with 106 aneurysms were identified in whom intraoperative flow measurements were available for analysis. ", "We assessed the frequency of flow compromise and clip repositioning and correlated these with postoperative angiography and stroke. ", "Significant (>25%) reduction in flow rate was apparent in 33 (31.1%) cases, and resulted in clip repositioning in 27 (25.5%), with return to baseline flow except for two cases with vessel thrombosis/dissection. ", "In the other six cases, flow reduction was owing to spasm resolving with papaverine (n = 3) or responded to retractor repositioning (n = 3). ", "In another six (5.7%) cases, unnecessary clip repositioning was avoided (n = 3) or safe occlusion of the parent vessel for trapping of the aneurysm was allowed by confirming adequate distal flow (n = 3). ", "Aneurysms of the basilar, middle cerebral, anterior communicating, or carotid terminus were more likely to be associated with flow compromise (odds ratio, 4.3; P = 0.03). ", "Postoperative angiography corroborated vessel patency in all cases, and no unexpected large vessel occlusions or strokes were evident. ", "Use of the ultrasonic flow probe provides real-time immediate feedback concerning vessel patency. ", "Vessel compromise is easier to interpret than with Doppler, and faster/less invasive than intraoperative angiography. ", "Intraoperative flow measurement is a valuable adjunct for enhancing the safety of aneurysm surgery." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0.016853932584269662, 0, 0, 0, 0, 0, 0, 0, 0, 0.00847457627118644, 0 ]

[ "Polyethylene glycol molecular weight influences the ClearT2 optical clearing method for spheroids imaging by confocal laser scanning microscopy.", "\nSome fluorescence microscopy techniques, such as confocal laser scanning microscopy (CLSM), have a limited penetration depth. ", "Consequently, the visualization and imaging of three-dimensional (3-D) cell cultures, such as spheroids, using these methods can be a significant challenge. ", "Therefore, to improve the imaging of 3-D tissues, optical clearing methods have been optimized to render transparency to the opaque spheroids. ", "The influence of the polyethylene glycol (PEG) molecular weight (MW) used in the ClearT2 method for the imaging of propidium iodide (PI)-stained spheroids was investigated. ", "The results demonstrated that the ClearT2 clearing method contributes to spheroids transparency and to the preservation of PI fluorescence intensity for all the PEG MW used (4000, 8000, and 10,000 Da). ", "Furthermore, the ClearT2 method performed using PEG 4000 Da allowed a better PI signal penetration depth and cross-section depth. ", "Overall, the optimization of PEG MW can improve the imaging of intact spheroids by CLSM. ", "Furthermore, this work may also contribute to increase the application of 3-D cell culture models by the pharmaceutical industry for the high-throughput screening of therapeutics." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.006944444444444444, 0, 0, 0, 0.011560693641618497, 0.01485148514851485, 0.015384615384615385, 0.011235955056179775, 0 ]

[ "Thor’s Anthony Hopkins: “Actors Are Pretty Stupid. ", "My Opinion Is Not Worth Anything”\n\nThor’s Anthony Hopkins: “Actors Are Pretty Stupid. ", "My Opinion Is Not Worth Anything”\n\nOscar-winning actor and star of films such as The Silence of the Lambs and Thor Anthony Hopkins discussed a number of topics with fellow actor Brad Pitt.", "\n\nIn an interview with Brad Pitt for Interview Magazine, Hopkins brought up the the subject of people asking him about “present situations.”", "\n\nHe stated:\n\n“People ask me questions about present situations in life, and I say, “I don’t know, I’m just an actor. ", "I don’t have any opinions.”", "\n\nRelated: Opinion: The Real Reason Doctor Sleep Flopped and Its Connection to Ewan McGregor’s Birds Of Prey Role\n\nHe added:\n\n“Actors are pretty stupid. ", "My opinion is not worth anything. ", "There’s no controversy for me, so don’t engage me in it, because I’m not going to participate.”", "\n\nPitt echoed Hopkins comments:\n\n“I feel the same. ", "I don’t care. ", "I’m glad things are always evolving and changing, and complaining about it won’t do any good. ", "You work with what you have. “", "\n\nHe continued:\n\n“Like digital, they’re doing amazing things with it, so I don’t lament that film is being used less and less.”", "\n\nRelated: Ad Astra Actor Brad Pitt: “President Trump Represents a Much Bigger Threat on Such More Serious Issues”\n\nPitt would then ask Hopkins, “What about the responsibility you feel when playing a real-life character? ", "There’s no way to get the actual life, beat-by-beat, of a person, but you aim for an essence, correct?”", "\n\nHopkins answered by discussing his role as Richard Nixon in the 1995 film Nixon from director Oliver Stone.", "\n\n“Oliver Stone gave me the part of Nixon, and I remember thinking, “Why would he give me that part?” ", "And he said, “Because I’ve read interviews about you being a loner. ", "That was Nixon.””", "\n\nHe continued:\n\n“So I watched a lot of Nixon films. ", "I went down to Yorba Linda, California, to see the house where he was born. ", "Bill Clinton told me that when he became president, he would phone Nixon every week.”", "\n\nRelated: Men in Black Star Emma Thompson Predicts Climate Crisis Will Force People To Eat Their Pets\n\nHopkins continued his story:\n\n“He’d ask him about China and Russia. ", "Clinton said this of him: He was a brilliant politician, but there was something in him that was so insecure. ", "And Oliver just tried to portray the man as he was—neither good or bad, but a man who makes mistakes, as we all do.”", "\n\nHe then details how emotional it was to play Nixon.", "\n\n“I found it very emotional to play him, because I could feel what it must have been like, the disgrace of having to resign. ", "And then the humiliation of having to say goodbye.”", "\n\nHe added:\n\n“And then I read about his wife. ", "She couldn’t bear to look down at the lawn of the White House as they took off.”", "\n\nStar Wars Actor Daisy Ridley: “Every Sane Person” Has an Issue With President Donald Trump\n\nPitt then comments on the film and Nixon’s life describing it as a “fascinating story from a humanistic point of view.” ", "He specifically points to the scene where Hopkins as Nixon asks, “Why don’t they like me?”", "\n\nHopkins responds to the comment saying:\n\n“That’s it. ", "You can see the pain in him, and you think, “Well, am I better than him? ", "No. ", "I’m not better than him. ", "I’ve got my own immoral quirks.””", "\n\nHopkins then recounts his drinking habit and the power of forgiveness.", "\n\n“Why do we do the things we do? ", "We have no idea. ", "I don’t know why I drank all my life. ", "I did it because it was the only thing I knew. ", "I look back now and I think, “Well, it wasn’t bad, but I don’t want to do it again.” ", "I caused a bit of damage then, but I’ve made my apologies to people for doing what I did. ", "It’s all part of being alive. ", "And whatever power there is in us, it’s forgiving. ", "It’s over. ", "Forget it. ", "Just move on.”", "\n\nThe actor, though Welsh-born became a naturalized U.S. Citizen back in 2000, hasn’t been active in political life as a new citizen.", "\n\nX-Men and James Bond Actor Alan Cumming: Having An Anti-Immigration Stance is Racist\n\nIn an interview with The Guardian in 2018, Hopkins stated, “I don’t vote because I don’t trust anyone. ", "We’ve never got it right, human beings. ", "We are all a mess, and we’re very early in our evolution.”", "\n\nHe then provides a brief history lesson.", "\n\n“Look back throughout history: you have the 20th century, the murder of 100 million people, barely 80 years ago. ", "The 1914-18 war, the civil war in America, slaughter, bloodshed… I don’t know if there’s a design in it, but it is extraordinary to look at it and get a perspective. ", "I think, ‘Well, if it’s the end, there’s nothing we can do about it, and it’ll blow over, whatever happens.’”", "\n\nWhile he might not be involved politically he did have some opinions about money and corporate power.", "\n\n“In America, they’re obsessed with healthy food. ", "They tell you, if you eat junk food, you get fat and you die. ", "Well, television is run by money and corporate power and sponsorship. ", "It’s junk food for the brain. ", "Toxic.”", "\n\nWhat do you think of Sir Anthony Hopkins’ ideas when it comes to actors and their opinions? ", "Let us know!", "\n\n(Visited 17,092 times, 1 visits today)" ]

{ "pile_set_name": "OpenWebText2" }

[ 0.0196078431372549, 0.011627906976744186, 0.015957446808510637, 0.014285714285714285, 0, 0, 0, 0, 0, 0.0196078431372549, 0, 0, 0, 0, 0.01809954751131222, 0, 0.027522935779816515, 0.0196078431372549, 0, 0.058823529411764705, 0.018867924528301886, 0.013157894736842105, 0.023529411764705882, 0, 0.00909090909090909, 0.008620689655172414, 0.018867924528301886, 0, 0, 0, 0.0125, 0.014018691588785047, 0.022222222222222223, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.007518796992481203, 0.020942408376963352, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.010638297872340425, 0, 0 ]

[ "Q:\n\nException on accessing a class in Web API\n\nI am encountering a weird behavior. ", "My Web API suddenly cannot be used because of this exception:\n[ArgumentNullException: Value cannot be null.", "\nParameter name: String]\n System.", "Number.", "StringToNumber(String str, NumberStyles options, NumberBuffer& number, NumberFormatInfo info, Boolean parseDecimal) +14198596\n System.", "Number.", "ParseInt32(String s, NumberStyles style, NumberFormatInfo info) +177\n Dota2IoT.Server.", "Models.", "ViewModels.", "NatMappingManager..cctor() +112\n\n[TypeInitializationException: The type initializer for 'Dota2IoT.Server.", "Models.", "ViewModels.", "NatMappingManager' threw an exception.]", "\n Dota2IoT.Server.", "MvcApplication.", "Application_Start() +223\n\nAnd this is the constructor of NapMappingManager:\n private ConcurrentDictionary<int, NatMappingViewModel> mappings;\n\n private NatMappingManager()\n {\n this.mappings = new ConcurrentDictionary<int, NatMappingViewModel>();\n }\n\nPreviously, I didn't even call NapMappingManager from the Global.asax Application_Start, but I think it may be because of some concurrent. ", "However, now I try to call:\n// Initialize the NAT\nModels.", "ViewModels.", "NatMappingManager.", "Instance.", "ToString()\n\nSo the constructor can be called, but the same error happens. ", "What's wrong with it? ", "Did I do something with the code, or a bug from ConcurrentDictionary?", "\nEDIT: The exception does not happen on my local machine. ", "It only happens on my Azure VM (Window Server 2012) server.", "\nEDIT to clarify for the correct answer:\nThis is the code that causes the problem (I do not have explicit static constructor):\n private static readonly int MaxDeviceNumber = int.", "Parse(ConfigurationManager.", "AppSettings[\"MaxDeviceNumber\"]);\n private static readonly int MinDeviceNumber = int.", "Parse(ConfigurationManager.", "AppSettings[\"MinDeviceNumber\"]);\n\nA:\n\n.cctor is the static constructor. ", "This fits with the fact that you have a type initializer exception. ", "It looks like you are trying to parse an int in the static constructor of NatMappingManager, but the value you are passing to it is null.", "\n(If you don't have an explicit static constructor, look for a static member that is being initialized on the same line that it is declared.)", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0, 0.014705882352941176, 0, 0.011363636363636364, 0, 0, 0.009523809523809525, 0, 0, 0, 0, 0.06666666666666667, 0.004901960784313725, 0.017543859649122806, 0, 0, 0, 0, 0, 0.014492753623188406, 0, 0, 0, 0, 0, 0, 0, 0, 0.0072992700729927005, 0, 0 ]

[ "Dave Urbanski, The Blaze, October 20, 2014\n\nFerguson protesters and football fans fought in the street after the St. Louis Rams game against the Seattle Seahawks on Sunday afternoon, the St. Louis Post-Dispatch reported.", "\n\nThe protesters were outside the Edward Jones Dome in St. Louis yelling and chanting about the killing of Michael Brown, the black 18-year-old who was fatally shot by a white police officer from Ferguson, Missouri, in August.", "\n\n{snip}\n\nThere was a spitting incident caught on video–but this one involved a female protester spitting in a man’s face.", "\n\nThe black female was heard on the video yelling a string of racially charged expletives at police and fans.", "\n\n“F*** all you white motherf***ers!” ", "she was heard yelling at one point.", "\n\nA white man was seen arguing with the woman as they stood almost toe-to-toe at one point.", "\n\nShe then backed out of frame–but the sound of spitting was heard and a visible clump of spit came at the man’s face.", "\n\n{snip}\n\nThe female was then surrounded by several police officers who placed her in handcuffs.", "\n\n{snip}\n\nPolice reported that two women, one of whom was a juvenile, were arrested after the street fight, the Post-Dispatch reported.", "\n\nMore from the Post-Dispatch concerning the brawl:\n\nAt one point during the scuffle, a protester carrying an upside-down American flag used the pole as a weapon, attempting to jab a fan with the pole. ", "A short time later, after the initial confrontation, a different fan grabbed the flag and ran with it for a brief period; protesters then wrestled the flag back. ", "Two people were injured including one woman who was punched in the face after chanting “Let’s go Rams” to protesters. ", "She was later treated and released by paramedics. ", "Police did not provide details on the second victim’s injuries. ", "Police said two groups of about 20 to 25 people each had argued and fought.", "\n\n{snip}\n\n[Editor’s Note: There is video of white fans being attacked by black protesters at the original article link below.]" ]

{ "pile_set_name": "OpenWebText2" }

[ 0.022727272727272728, 0.008849557522123894, 0, 0, 0, 0, 0, 0, 0, 0.007407407407407408, 0.0049504950495049506, 0, 0.00847457627118644, 0, 0, 0, 0 ]

[ "Malignant mediastinal lymphadenopathy detected by endoscopic ultrasound and guided fine needle aspiration in patients with resectable pancreaticobiliary cancer.", "\nResection surgery for pancreaticobiliary malignancies carries significant morbidity and mortality. ", "Hence, preoperative assessment to exclude unresectable disease is mandatory. ", "CT abdomen is the primary modality for staging of pancreaticobiliary cancers. ", "However, some patients have malignant mediastinal lymphadenopathy (MML), which may be detected on endoscopic ultrasound (EUS) but not on CT scan. ", "We prospectively evaluated 75 consecutive patients (median age 54 years: 44 men) with a diagnosis of resectable pancreaticobiliary cancer (carcinoma gallbladder, carcinoma pancreas, cholangiocarcinoma, or periampullary carcinoma) for the presence of MML using EUS by an experienced endosonographer. ", "If a lymph node had one or more features suggestive of malignancy, i.e. size exceeding 1 cm, hypoechoic appearance, a round shape, and regular margins, it was subjected to EUS-FNA. ", "In seven (9.3%; 95% confidence intervals: 3.8% to 18.2%) of the 75 patients, EUS revealed enlarged mediastinal lymph nodes. ", "The location of these lymph nodes was subcarinal in three, paraesophageal in two, and paratracheal in one patient; another patient had lymph nodes at two sites, i.e. the subcarinal and aortopulmonary window. ", "In four of these seven patients, FNA documented the presence of MML. ", "The overall rate of pathologically proven MML was 4/75 (5.3%; 95% CI [1.4% to 13%]). ", "EUS-FNA diagnosed MML in 5.3% of patients with pancreaticobiliary cancer. ", "It may be useful to consider EUS assessment in patients with otherwise resectable pancreaticobiliary malignancy." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0, 0.00684931506849315, 0.0033444816053511705, 0.0055248618784530384, 0.008064516129032258, 0, 0.014492753623188406, 0.011764705882352941, 0.013513513513513514, 0.008928571428571428 ]

[ "Marko Strahija\n\nMarko Strahija (born May 28, 1975 in Zagreb) is a backstroke swimmer from Croatia, who competed at three Summer Olympics between 1996 and 2008.", "\n\nMarko attended The Ohio State University and held university records in the 100 and 200 backstroke before the times were eclipsed by RJ Lemyre.", "\n\nIn Atlanta, United States, he finished in 11th position (B-Final) in the Men's 200 m backstroke, and was eliminated in the heats of the Men's 4 × 200 m freestyle relay, alongside Gordan Kožulj, Miroslav Vučetić, and Marijan Kanjer. ", "Strahija won the silver medal, behind Aaron Peirsol of the United States, in the 200 m backstroke at the 2002 FINA Short Course World Championships.", "\n\nDoping suspension and exoneration\nStrahija's urine sample, taken in an out-of-competition control in July 2002, tested positive for human chorionic gonadotropin (hCG). ", "Since hCG is a known tumor marker, Strahija underwent medical tests that found nothing suspicious. ", "He maintained his innocence and disputed scientific validity of testing for hCG. ", "Nevertheless, in 2003 he received a two-year suspension.", "\n\nIn October 2007 Strahija again tested positive for hCG, which caused him to miss the European Short Course Championships held in December that year. ", "This time, subsequent medical tests found testicular cancer, and Strahija immediately underwent surgery. ", "In February 2008, the International Swimming Federation exonerated Strahija, lifting his provisional suspension. ", "He made a successful recovery, taking part in the 2008 Summer Olympics.", "\n\nOlympic results\n\nReferences\n\nExternal links\n\n Short profile on Croatian Olympic Committee \n\nCategory:1975 births\nCategory:Living people\nCategory:Male backstroke swimmers\nCategory:Croatian male swimmers\nCategory:Olympic swimmers of Croatia\nCategory:Swimmers at the 1996 Summer Olympics\nCategory:Swimmers at the 2000 Summer Olympics\nCategory:Swimmers at the 2008 Summer Olympics\nCategory:Sportspeople from Zagreb\nCategory:Testicular cancer survivors\nCategory:Croatian sportspeople in doping cases\nCategory:Doping cases in swimming\nCategory:Medalists at the FINA World Swimming Championships (25 m)\nCategory:European Aquatics Championships medalists in swimming\nCategory:Mediterranean Games gold medalists for Croatia\nCategory:Mediterranean Games silver medalists for Croatia\nCategory:Competitors at the 2001 Mediterranean Games\nCategory:Mediterranean Games medalists in swimming" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0, 0.013793103448275862, 0.01282051282051282, 0.006756756756756757, 0, 0.010101010101010102, 0.012345679012345678, 0, 0.006622516556291391, 0.009523809523809525, 0.017699115044247787, 0, 0.00683371298405467 ]

[ "Children's access to mental health care: does insurance matter?", "\nUsing data from a 1992 community survey of children and their parents (or guardians), we found major gaps in mental health insurance coverage. ", "Interestingly, private insurance had no statistically significant effect on use of mental health services. ", "Youth without insurance coverage and those with public insurance had higher rates of serious emotional disorder than did those with private insurance. ", "The analysis is based on the National Institute of Mental Health's Methods for the Epidemiology of Child and Adolescent Mental Disorders (MECA) study, conducted in three mainland U.S. sites and in Puerto Rico." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0, 0.004784688995215311 ]

[ "Neyzar, Qom\n\nNeyzar (, also Romanized as Neyzār, Nayzar, and Nāzar) is a village in Neyzar Rural District, Salafchegan District, Qom County, Qom Province, Iran. ", "At the 2006 census, its population was 320, in 91 families.", "\n\nReferences \n\nCategory:Populated places in Qom Province" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0, 0, 0 ]

[ "Q:\n\nUpgrading starter wire?", "\n\nI did an alternator upgrade recently, and I have 3 feet of 1/0awg wire left over. ", " Would it make any difference if I were to use this remaining wire to replace my starter's current 6awg wire?", "\nI have a 4.0L jeep wrangler and I have to crank it for a couple seconds to get it to start. ", " Would adding a wire with lower resistance help anything? ", " I don't see how I could hurt anything by adding 1/0awg wire, but would it help at all?", "\n\nA:\n\nIt certainly shouldn't hurt anything, but it is unlikely to help either as the wires to your starter should be appropriate anyway. ", "\nIf your wiring is old then you may see an improvement, otherwise I would suggest leaving it alone.", "\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "2019 Bangkok Challenger II\n\nThe 2019 Bangkok Challenger II was a professional tennis tournament played on hard courts. ", "It was part of the 2019 ATP Challenger Tour. ", "It took place in Bangkok, Thailand between 18 and 24 February 2019.", "\n\nSingles main draw entrants\n\nSeeds\n\n 1 Rankings are as of 11 February 2019.", "\n\nOther Entrants\nThe following players received wildcards into the singles main draw:\n Thanapet Chanta\n Pruchya Isaro\n Palaphoom Kovapitukted\n Janko Tipsarević\n Wishaya Trongcharoenchaikul\n\nThe following players received entry into the singles main draw using their ITF World Tennis Ranking:\n Dimitar Kuzmanov\n David Pérez Sanz\n Oriol Roca Batalla\n Denis Yevseyev\n \nThe following players received entry from the qualifying draw:\n Bai Yan\n Ivan Gakhov\n\nThe following player received entry as a lucky loser:\n Moez Echargui\n\nChampions\n\nSingles\n\n James Duckworth def. ", " Alejandro Davidovich Fokina 6–4, 6–3.", "\n\nDoubles\n\n Li Zhe / Gonçalo Oliveira def. ", " Enrique López Pérez / Hiroki Moriya 6–2, 6–1.", "\n\nBangkok Challenger II\n2019\nCategory:2019 in Thai sport\nCategory:February 2019 sports events in Asia" ]

{ "pile_set_name": "Wikipedia (en)" }

[ 0, 0, 0, 0, 0.008665511265164644, 0, 0.022222222222222223, 0.0425531914893617, 0 ]

[ "Arsenal open talks with agent of £23m-rated Brazilian starlet\n\nAccording to a report in the Express, Arsenal have opened talks with the representatives of Sao Paulo midfielder Marquinhos Cipriano about a potential move to the Emirates this summer.", "\n\nMarquinhos was born in Catanduva – a small town just outside of Sao Paulo – and has developed in to one of Brazil’s hottest young talents since joining Sao Paulo in September 2015.", "\n\nThe 17-year-old has averaged more than a goal a game so far after scoring 14 goals in just 12 appearances for the Sao Paulo Under-17 team and coaches at the club feel he has what it takes to break in to the first team in the near future.", "\n\nIt looks like Marquinhos’ potential has caught the attention of some of Europe’s most powerful clubs as the Express are citing a report coming from ESPN Brazil that claims Arsenal have made contact with the players agent to discuss a potential move to North London.", "\n\nArsene Wenger is always on the look-out for the best young talent in the world so it’s no surprise to hear he’s hot on the trail of Marquinhos, but any deal isn’t going to be easy to pull off.", "\n\nFirstly, the Express claim that Sao Paulo are determined to keep hold of the teenager and will refuse to sell for anything below his buyout fee – which is set at an eye-watering £23m.\n\nI think there is virtually no chance Wenger will spend that kind of money on a 17-year-old who’s not even made a single appearance for a senior team yet so Sao Paulo are going to have to lower their demands considerably.", "\n\nHowever, even if we do manage to secure a deal, we aren’t the only club in the running as the report suggests that Italian champions Juventus have also been in touch with the midfielders representatives about a move to Serie A." ]

{ "pile_set_name": "Pile-CC" }

[ 0.004048582995951417, 0, 0, 0.011235955056179775, 0.005154639175257732, 0.002457002457002457, 0.004366812227074236 ]

[ "Effect of deposit age on adsorption and desorption behaviors of ammonia nitrogen on municipal solid waste.", "\nAmmonia nitrogen pollution control is an urgent issue of landfill. ", "This research aims to select an optimal refuse for ammonia nitrogen removal in landfill from the point of view of adsorption and desorption behavior. ", "MSW (municipal solid waste) samples which deposit ages were in the range of 5 to 15 years (named as R(15), R(11), R(7), and R(5)) were collected from real landfill site. ", "The ammonia nitrogen adsorption behaviors of MSW including equilibrium time, adsorption isotherms, and desorption behaviors including equilibrium time were determined. ", "Furthermore, the effects of pH, OM, Cu(II), Zn(II), and Pb(II) on adsorption and desorption behavior of ammonia nitrogen were conducted by orthogonal experiment. ", "The equilibrium time of ammonia nitrogen adsorption by each tested MSW was very short, i.e., 20 min, whereas desorption process needed 24 h and the ammonia nitrogen released from refuses was much lesser than that adsorbed, i.e., accounted for 3.20 % (R(15)), 14.32 % (R(11)), 20.59 % (R(7)), and 20.50 % (R(5)) of each adsorption quantity, respectively. ", "The maximum adsorption capacity estimated from Langmuir isotherm appeared in R(15)-KCl, i.e., 25,000 mg kg(-1). ", "The best condition for ammonia nitrogen removal from leachate was pH >7.5, OM 23.58 %, Cu(II) <5 mg L(-1), Zn(II) <10 mg L(-1), and Pb(II) <1 mg L(-1). ", "Ammonia nitrogen in landfill leachate could be quickly and largely absorbed by MSW but slowly and infrequently released. ", "The refuse deposited for 15 years could be a suitable material for ammonia nitrogen removal." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0.011764705882352941, 0.005952380952380952, 0.006172839506172839, 0.005649717514124294, 0.017857142857142856, 0, 0, 0 ]

[ "The Black Panther trailer feels like a breath of fresh air in an environment crowded with superhero movies – no comic book adaptation has ever looked like this. ", "Heck, no movie has ever looked like this. ", "Even with a few familiar Marvel Studios trappings on display, Ryan Coogler’s movie looks to blend superheroes and afrofuturism and all kinds of intrigue into something…well, new.", "\n\nAnd if you’re new to this corner of the Marvel universe (or just want to take a closer look), we went through the trailer frame-by-frame for an extended breakdown. ", "Join us, won’t you?", "\n\nThe trailer opens with something we won’t see much in the rest of the trailer: white guys. ", "However, these two white guys should be familiar to fans of the Marvel Cinematic Universe. ", "On the left is Everett K. Ross, a CIA agent who was initially introduced in Captain America: Civil War. ", "On the right is Ulysses Klaue, a black market smuggler and arms dealer who first reared his loathsome head in Avengers: Age of Ultron. ", "Despite these characters being introduced elsewhere, both are commonly associated with the Black Panther world in the Marvel comics. ", "And this is not the first time actors Martin Freeman and Andy Serkis have shared a big scene together – they previously had a showdown in The Hobbit, where they played Bilbo Baggins and Gollum, respectively.", "\n\nThis shot offers our first glimpse at what makes the fictional land of Wakanda so very cool and special. ", "In most genre stories, wealth and technology seem to immediately westernize a nation of people. ", "But not here. ", "Instead, the secretive country of Wakanda is distinctly African while also being technologically advanced to the point of full-on science fiction.", "\n\nThis is best summed up by T’Challa having an audience on a waterfall with people in traditional robes and garb…after he was dropped off by his personal super-jet. ", "This is only the tip of the iceberg when it comes to the wild afrofuturism featured elsewhere in the trailer. ", "According to Entertainment Weekly, this is the Warrior Falls, where kings are coronated. ", "As you may remember, T’Challa’s father died in Civil War – this must be the ceremony to give him the throne.", "\n\nThese opening shots also introduce our hero in style. ", "First, we see some kind of heavily armed convoy making their way through the jungle. ", "It’s not clear who these people are, but Wakanda’s borders are closed, which surely makes them intruders. ", "They could be smugglers, looking to steal some of the nation’s precious vibranium (which is the source of Wakanda’s wealth and tech in addition to being the stuff in Captain America’s shield). ", "Could they be in league with Ulysses Klaue, who was smuggling vibranium in Age of Ultron?", "\n\nBlack Panther is revealed in the treetops by the glow of a flashlight, leading to some action beats that feels more like something out of a Batman movie than a Marvel movie. ", "T’Challa is much faster and much stronger than the average human (we’ll touch on that in a moment), but surprise and fear are two key weapons in his arsenal. ", "He makes quick work of these guys.", "\n\nMuch like how the post-Avengers MCU movies went out of their way to remind audiences that Captain America has literal super-strength and can kick cars like tin cans, this trailer reminds us that Black Panther is strong enough to fling a foe into a car so hard that it crumples.", "\n\nCivil War didn’t have time to stop and breathe and explain what the deal is with T’Challa and why he can stand toe-to-toe with the Avengers. ", "The new film will surely rectify that. ", "In the comics, the Black Panther mantle is passed down over generations and the current bearer of the title has a special connection with the Wakandan Panther God and is allowed to eat a special Heart-Shaped herb that increases everything from strength to agility to senses. ", "Despite the sci-fi tinge of Wakanda, there is a mystical streak running through the Black Panther world that could easily tie into the likes of Doctor Strange and Thor.", "\n\nAnd finally, we see who is watching Ross interrogate Klaue: T’Challa himself and Okoye, his chief of security played by The Walking Dead‘s Danai Gurira. ", "She’s a member of the Dora Milaje, a group that we will see frequently throughout the rest of the trailer.", "\n\nWelcome to Wakanda! ", "Cut off from the rest of the world and sitting on the world’s largest supply of vibranium, the nation’s capital puts every other city in the MCU to shame. ", "As we will see in other shots, the city has not lost its heritage in a maze of futuristic skyscrapers.", "\n\nUnderneath that paint is Academy Award winner Forrest Whitaker, playing Zuri, a Wakandan shaman and an important religious figure. ", "As he is present for when T’Challa arrives on Warrior Falls, we can assume that he is the one directly in charge of the new king’s coronation.", "\n\nWe’ll see more of him soon, but the man under this mask is Erik Killmonger, played by Ryan Coogler regular Michael B. Jordan. ", "We’ll talk more about him soon. ", "But once again, the combination of his futuristic armor and tribal mask is unlike anything we’ve seen before in the MCU. ", "In fact, the mask is a reference to his costume in the comics (but don’t expect him to wear it too often in the film).", "\n\nAnd he has a gang. ", "And some explosives. ", "And that graffiti on the wall suggests that we are definitely not in Wakanda." ]

{ "pile_set_name": "OpenWebText2" }

[ 0.006211180124223602, 0, 0.011235955056179775, 0.006024096385542169, 0, 0, 0.01098901098901099, 0.019230769230769232, 0.022222222222222223, 0.015037593984962405, 0.01932367149758454, 0.009345794392523364, 0, 0, 0, 0, 0, 0.02247191011235955, 0, 0, 0, 0.009433962264150943, 0.0051813471502590676, 0.02247191011235955, 0.017045454545454544, 0, 0, 0.0035842293906810036, 0, 0, 0.0036363636363636364, 0.011904761904761904, 0.012903225806451613, 0, 0.045454545454545456, 0.0064516129032258064, 0, 0.03007518796992481, 0, 0.0234375, 0, 0.008264462809917356, 0, 0, 0, 0 ]

[ "/* ----------------------------------------------------------------------------\n * This file was automatically generated by SWIG (http://www.swig.org).", "\n * Version 3.0.11\n *\n * Do not make changes to this file unless you know what you are doing--modify\n * the SWIG interface file instead.", "\n * ----------------------------------------------------------------------------- */\n\npackage com.badlogic.gdx.physics.bullet.extras;\n\nimport com.badlogic.gdx.physics.bullet.", "BulletBase;\nimport com.badlogic.gdx.physics.bullet.linearmath.*;", "\nimport com.badlogic.gdx.physics.bullet.collision.*;", "\nimport com.badlogic.gdx.physics.bullet.dynamics.*;", "\nimport com.badlogic.gdx.math.", "Vector3;\nimport com.badlogic.gdx.math.", "Quaternion;\nimport com.badlogic.gdx.math.", "Matrix3;\nimport com.badlogic.gdx.math.", "Matrix4;\nimport com.badlogic.gdx.physics.bullet.inversedynamics.", "MultiBodyTree;\nimport com.badlogic.gdx.physics.bullet.dynamics.btDynamicsWorld;\nimport com.badlogic.gdx.physics.bullet.dynamics.btContactSolverInfo;\nimport com.badlogic.gdx.physics.bullet.collision.btCollisionShape;\n\npublic class User2InternalIndex extends BulletBase {\n\tprivate long swigCPtr;\n\t\n\tprotected User2InternalIndex(final String className, long cPtr, boolean cMemoryOwn) {\n\t\tsuper(className, cPtr, cMemoryOwn);\n\t\tswigCPtr = cPtr;\n\t}\n\t\n\t/** Construct a new User2InternalIndex, normally you should not need this constructor it's intended for low-level usage. */ ", "\n\tpublic User2InternalIndex(long cPtr, boolean cMemoryOwn) {\n\t\tthis(\"User2InternalIndex\", cPtr, cMemoryOwn);\n\t\tconstruct();\n\t}\n\t\n\t@Override\n\tprotected void reset(long cPtr, boolean cMemoryOwn) {\n\t\tif (!", "destroyed)\n\t\t\tdestroy();\n\t\tsuper.reset(swigCPtr = cPtr, cMemoryOwn);\n\t}\n\t\n\tpublic static long getCPtr(User2InternalIndex obj) {\n\t\treturn (obj == null) ? ", "0 : obj.swigCPtr;\n\t}\n\n\t@Override\n\tprotected void finalize() throws Throwable {\n\t\tif (!", "destroyed)\n\t\t\tdestroy();\n\t\tsuper.finalize();\n\t}\n\n @Override protected synchronized void delete() {\n\t\tif (swigCPtr !", "= 0) {\n\t\t\tif (swigCMemOwn) {\n\t\t\t\tswigCMemOwn = false;\n\t\t\t\tExtrasJNI.delete_User2InternalIndex(swigCPtr);\n\t\t\t}\n\t\t\tswigCPtr = 0;\n\t\t}\n\t\tsuper.delete();\n\t}\n\n public User2InternalIndex() {\n this(ExtrasJNI.new_User2InternalIndex(), true);\n }\n\n public void addBody(int body, int parent) {\n ExtrasJNI.User2InternalIndex_addBody(swigCPtr, this, body, parent);\n }\n\n public int buildMapping() {\n return ExtrasJNI.User2InternalIndex_buildMapping(swigCPtr, this);\n }\n\n public int user2internal(int user, java.nio.", "IntBuffer internal) {\n assert internal.isDirect() : \"Buffer must be allocated direct.\";", "\n {\n return ExtrasJNI.User2InternalIndex_user2internal(swigCPtr, this, user, internal);\n }\n }\n\n public int internal2user(int internal, java.nio.", "IntBuffer user) {\n assert user.isDirect() : \"Buffer must be allocated direct.\";", "\n {\n return ExtrasJNI.User2InternalIndex_internal2user(swigCPtr, this, internal, user);\n }\n }\n\n}\n" ]

{ "pile_set_name": "Github" }

[ 0.013245033112582781, 0, 0.011494252873563218, 0, 0, 0, 0, 0.02631578947368421, 0, 0, 0, 0.012280701754385965, 0.0297029702970297, 0.013071895424836602, 0.011627906976744186, 0.008620689655172414, 0, 0.011111111111111112, 0, 0.012195121951219513, 0 ]

[ "Q:\n\nPHP Web Design: Using Smarty vs Rest feeds with jQuery\n\nI've been trawling SO for a couple of hours tonight looking for some answers, but I've not found anything that really answers what I'm after. ", "My apologies if it has, in fact, been answered already.", "\nI'm designing a new website and I'm trying to decide on the architecture to use to serve the content. ", "In the past, my websites have used PHP feeding data into Smarty templates.", "\nHowever, recently at my work, I've been working on a Java web application where jQuery was used to retrieve the data from a RESTful API (which returned JSON), where HTML template pages were used as the base and javascript was used (utilising jQuery) to fill in the content.", "\nMy question: \nThe website I'm designing will be in PHP, but would it be better to construct (or use an existing) RESTful API or to continue on as I've done before feeding the data into Smarty templates?", "\nAre there real benefits to one or the other, or does it just come down to developer preference/experience?", "\nIf it helps, the website will be for a church, where the content types will be CMS-like; news/announcements, wiki-like pages, and a limited type of social networking (for the minister to communicate with parishioners).", "\n\nA:\n\nShort answer: It sounds like filling the content with JavaScript would not be useful in your case. ", "Loading the data with JavaScript is adding a layer of complexity with minimal or no benefit (in your case). ", "Take a look at CMSs and websites that have similar functionality to what you're doing. ", "WordPress, Drupal, etc.", "\nFor an example of when it might be useful to load data with JavaScript, check out the tags section on this site. ", "When you search for a tag it queries the server without reloading the page. ", "However, the initial tag list is loaded during the initial page load without any JS.", "\nHere are some cases where you might benefit from loading information with JS:\n\nThere is some data that will take longer to load. ", "For example, you're getting data from a web service. ", "Using the traditional method, you would need all the data you need for the page to be available before it can be displayed. ", "If you loaded that data with JavaScript, the page could load and the slow data would appear when it returns. ", "Realistically, you would probably just cache the data but this is just an example.", "\nYou will be getting more data as the user stays on the page. ", "You might want to update the page without refreshing.", "\nYou want the user to be able to query more data without refreshing the page.", "\nYour users have limited bandwidth (mobile).", "\nThis site has more guidelines: http://www.oracle.com/technetwork/articles/javase/index-137171.html\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0.0049504950495049506, 0, 0, 0.013513513513513514, 0.014598540145985401, 0.009852216748768473, 0, 0.0045662100456621, 0.009523809523809525, 0.009259259259259259, 0, 0.043478260869565216, 0.008771929824561403, 0, 0, 0, 0, 0, 0.009174311926605505, 0, 0, 0, 0, 0, 0.009900990099009901 ]

[ "Look Up Another Term\n\nDefinition: XRDS\n\n(EXtensible Resource Descriptor Sequence) An XML-based document that is the authoritative source of data about a registered i-name. ", "On behalf of XDI.org, the XRDS location is maintained by .us domain registrar NeuStar, Inc. at www.xri.net. ", "The XRDS, which contains the i-name, i-number and URLs to all services associated with that i-name, is updated through an i-broker. ", "See xri.net, i-broker, i-name and XRI." ]

{ "pile_set_name": "Pile-CC" }

[ 0.005813953488372093, 0.027777777777777776, 0.007575757575757576, 0.02631578947368421 ]

[ "Realtor\n\nFind Realtor In Weatherford TX\n\nYou should get a professional seller real estate agent with 5-star reviews. ", "Call a distinguished seller real estate agent with 5-star ratings today! ", "★★★★★\n\nMy uncomplicated focus is on the bottom line for the homeowner which is to offer the home for the greatest income within the time frame planned.", "\nOne factor of why I enjoy being a homeowner real estate agent is related to the fact that it permits me to make use of a variety of capabilities that I have actually created for many years to achieve a premium result for my clients. ", "These skills feature imagination, remodeling expertise, and regional property market conditions and property trends.", "\nFor me that is actually a much more favorable career decision being actually a homeowner agent due to my certain abilities.", "\nI have the product, the passion, the patience, the perseverance and, most importantly, the persuasive, some might even say, pervasive, personality. ", "My God-given abilities and expertise are actually naturally fit well to positive negotiations.", "\n\nThe buyer broker is actually usually a negative negotiator, making due dates, deadlines, and demands and divulging discrepancies, defects, and deficiencies as well as displaying a deadpan face, regularly acting depressed and dismayed regarding the residence of the seller, as well as demonstrating defeat, disenchantment, despair, disillusionment, dissatisfaction, and disappointment.", "\nCertainly, this is an oversimplified description but I have truly seen it played out a lot of times. ", "I possess passion concerning the product I am selling: the residence is special, it is out-of-the-ordinary. ", "I sound like, as well as really strongly believe that, I bargain from power, never from a position of weakness.", "\nI understand the best ways to negotiate and also constantly feel I can easily out-negotiate a purchaser real estate agent considering that I really have the higher ground with the sun behind me. ", "Mine is a peaceful form of negotiating whereas the purchaser real estate broker has to resort to depressive, damaging, negotiating approaches.", "\n\nThe purchaser real estate broker pout of disappointment is actually, typically, the very first indicator that I am winning in the negotiations. ", "I get to utilize my natural positive personality to be gracious to buyer real estate agents.", "\nDebbie\nCopeland\nReal\nEstate\nAgent\nWeatherford\nTX\nPhone\n:(682)\n803-2153\nBusiness\nHours:\nOpen\n24/7" ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0, 0.006622516556291391, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.020618556701030927 ]

[ "Prognostic implication of dual-phase PET in adenocarcinoma of the lung.", "\nThe purpose of this study was to determine whether dual-phase (18)F-FDG PET could be a prognostic factor for adenocarcinoma of the lung. ", "One hundred patients with histologically proven adenocarcinoma of the lung were included in this retrospective analysis. ", "All patients underwent dual-phase pretherapy (18)F-FDG PET for which, after the intravenous administration of (18)F-FDG, both early (approximately 60 min) and delayed (approximately 90 min) PET were acquired. ", "The percentage change in the maximal standardized uptake values (SUVmax) of the cancer between the early and the delayed images was calculated. ", "Overall survival of the SUVmax change over time together with the known prognostic factors were calculated by the Kaplan-Meier method and evaluated with the log-rank test. ", "The prognostic significance was assessed by univariate and multivariate analyses. ", "Statistical analysis showed that SUVmax change over time between the early and the delayed PET was a strong independent predictor of outcome for lung adenocarcinoma. ", "A cutoff of 25% change for SUVmax over time showed the best discriminative value. ", "Patients with more than 25% increase in SUVmax had a median survival of 15 mo, compared with 39 mo for those with less than 25% increase in SUVmax. ", "Dual-phase (18)F-FDG PET reflects the dynamics of glucose metabolism. ", "Our findings suggest that the percentage SUVmax change over time is a strong prognostic factor in patients with lung adenocarcinoma and can be complementary to the other well-known factors." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.014084507042253521, 0, 0, 0.004784688995215311, 0, 0, 0, 0.006024096385542169, 0, 0, 0.014285714285714285, 0 ]

[ "So first off I'd like to apologize for not being around much. ", "Apparently a few of you have missed me and I am sorry. ", "I have just been redonculously busy lately.", "\n\n\n\nAnd to make matters worse...today was shit.", "\n\n\n\nI had a lovely Easter. ", "All my family came over to my parents' and we had a great lil brunch get-together. ", "Afterward, my mom was driving me back to my apartment in the city. ", "She asked me what I thought of the new church they were going to (where we went that morning) and I sort of gave an ambivalent response. ", "She asked if I missed church at all, and I decided that I wasn't going to lie about this anymore. ", "So one question led to another and she finally asked:\n\n\n\n\"So what do you believe in?\"", "\n\n\n\nto which I responded\n\n\n\n\"I don't believe in god.\"", "\n\n\n\nShe freaked.", "\n\n\n\nShe started crying and yelling. ", "She wondered how I could just give up god after so long. ", "She didn't understand where I got my answers then. ", "She wants to know how I could let science get in the way of my faith (when I told her religion contradicted too much with scientific fact).", "\n\n\n\nShe told me...that she felt like a failure because all she ever wanted was to raise good christian children.", "\n\n\n\nWhat the fuck am I supposed to say to that?", "\n\n\n\nNevermind that I am a good girl and have never gotten into any serious trouble. ", "I do well in school and am responsible. ", "I am (I feel) a pretty good daughter. ", "Obviously none of that matters. ", "I feel that I could have told her I was pregnant and it would have been less of an issue as long as I still wuvved jeebus.", "\n\n\n\nIt went on a bit, then she dropped me off in awkward silence, saying \"this isn't the end of this.\"", "\n\n\n\nGreat. ", "Just fucking great.", "\n\n\n\nSo this wasn't a very meaty post, but I needed to vent. ", "And who better to share my coming out story with than all my thinkatheist pals.", "\n\n\n\nAnyone else with sticky outting stories have any advice for a very stressed out Sniz?" ]

{ "pile_set_name": "OpenWebText2" }

[ 0, 0, 0, 0, 0.037037037037037035, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.011235955056179775 ]

[ "Introduction\n============\n\nThe components of the cell wall of Gram-positive bacteria may vary considerably depending on the genus, species, and even the strain ([@B6]). ", "Hence, the diverse structures of lactobacillus surface molecules, in large part, reflect the adaptation of these microorganisms to specific ecological and environmental conditions.", "\n\nThe surface polysaccharides of lactobacilli contribute significantly to their interaction with the environment, including interactions with other microorganisms and a variety of hosts, in particular with the intestinal mucosa and IEC, contributing to the probiotic characteristics of a specific strain. ", "Bacterial polysaccharides vary in composition, size and the position of branches, contributing to the structural diversity of surface molecules ([@B29]). ", "The polysaccharide component of a bacterial cell's surface may form EPS that are weakly attached to the cell surface, as well as CPS usually covalently bound to peptidoglycan, forming a firm coating around the cell. ", "It should be noted that the composition, size, and surface properties of polysaccharides depend on the growth conditions of a given bacterial species, as was recently shown for the surface properties of *Lactobacillus acidophilus* NCC2628 ([@B47]). ", "Based on its chemical composition, an EPS may be a homopolysaccharide -- consisting of one type of monosaccharide -- or heteropolysaccharide -- composed of different types of monosaccharides. ", "Generally, heteropolysaccharides synthesized by lactic acid bacteria may differ in composition, structure, function, size, and molecular weight ([@B5]). ", "The work of [@B38] revealed huge diversity of EPS produced by 31 strains belonging to different species of lactic acid bacteria. ", "The molecular weight of the EPS varied between 8 to \\>5,000 kDa. ", "Interestingly, it was shown that low molecular weight-EPS are more common than high molecular weight-EPS in lactic acid bacteria ([@B38]). ", "For instance, it has been shown that the EPS of *L. casei* YIT 9018 is composed of two types of heteropolysaccharides ([@B40]).", "\n\nGenerally, the EPS biodiversity could be result of different sugar nucleotide biosynthesis pathways and genetic variability of the strains ([@B38]). ", "The genetic elements involved in the synthesis of the conserved structure of the polymer are organized into operons, located on chromosomal or plasmid DNA ([@B2]; [@B56]). ", "The genes present in the *eps* operons encode proteins with specific functions: assembling of repeating units, chain length determination, polymerization, export, and regulation ([@B5]). ", "A number of glycosyltransferases involved in the repeating units' assembly have been revealed ([@B38]). ", "Characterization of a large number of operons for the biosynthesis of surface polysaccharides has revealed that within the conserved structural organization, variations in the number and position of genes in the operon exist ([@B42]).", "\n\nDue to their long history of safe use in human consumption, lactobacilli have been given \"Qualified Presumption of Safety\" status ([@B9]). ", "They have been proposed as beneficial for human and animal health as probiotics, defined as \"live microorganisms which when administered in adequate amounts confer a health benefit on the host\" ([@B10]).", "\n\nThe literature data suggest that surface polysaccharides indirectly affect the adhesive properties of lactobacilli by means of various interactions such as passive forces, electrostatic interactions, and hydrophobic and steric forces ([@B49]). ", "The work of [@B41] showed that EPS-producing strain *L. paraplantarum* BGCG11 have lower adhesion to IEC than its non-EPS-producing derivative NB1. ", "The results of [@B44] showed that the adhesion levels depend on the EPS type. ", "For example, the EPS fraction GG showed no effect on the adhesion of *L. rhamnosus* GG, while at the highest dose it slightly increased the adherence of *Bifidobacterium animalis* IPLA-R1. ", "However, the EPS fractions NB667 and IPLA-R1 promoted a significant reduction in the adhesion of both probiotic strains *B. longum* NB667 and *B. animalis* IPLA-R1. ", "Further, [@B7] showed that deletion of the entire EPS cluster slightly increased the gut persistence time of the knock-out mutant in comparison to the parental strain. ", "In addition, the results of [@B43] indicated that polysaccharides produced by *L. rhamnosus* E/N hindered adhesion and aggregation by masking protein receptors at the cell surface. ", "Moreover, the reduction in EPS production by *L. johnsonii* FI9785 increased bacterial adhesion to chicken gut explants ([@B8]).", "\n\nLactobacillus strains *L. paraplantarum* BGCG11 and *L. fermentum* BGHI14 have been shown to have immunomodulatory capacity and to have the ability to induce the production of various cytokines (IL-17, IFN-γ, IL-10, TNF-α, and IL-1β; [@B41]; [@B34]). ", "The EPS-producing strain *L. paraplantarum* BGCG11, as an example, was shown to exhibit an anti-inflammatory and immunosuppressive profile, whereas its non-EPS derivative induced a pro-inflammatory response ([@B41]).", "\n\nInterest in the health-promoting potential of lactobacilli has significantly increased, especially with regard to their role in the prevention and treatment of disease. ", "Our previous results showed that the natural dairy isolate *L. paracasei* subsp. *", "paracasei* BGSJ2-8 synthesizes BacSJ bacteriocin, PI-type proteinase and AggLb aggregation promoting protein, and it is able to survive in simulated gut conditions, hence it could be used in dairy industry as functional starter culture with potential probiotic traits ([@B33]; [@B37]). ", "In this study, the effects of an EPS produced by BGSJ2-8 (EPS-SJ) on its adhesion to IEC, a decrease of *Escherichia coli* association with Caco-2 cells, and the proliferation of lymphocytes in GALT were investigated. ", "The knowledge obtained in this study may help to understand the specific contribution of EPS-SJ in gut colonization and pathogen exclusion.", "\n\nMaterials and Methods {#s1}\n=====================\n\nMedia and Growth Conditions\n---------------------------\n\nThe strains and plasmids used in this study are listed in **Table [1](#T1){ref-type=\"table\"}.** ", "Lactobacillus strains were grown in De Man-Rogosa-Sharpe (MRS) medium (Merck GmbH, Darmstadt, Germany) at 28°C, 30°C or 37°C when it was appropriate. *", "E. coli* strains, used as intermediate hosts for cloning experiments and plasmid propagation, were grown in LBwith aeration at 37°C for 16 h. Solid media were prepared by adding agar (1.5% w/v). ", "When appropriate, the media contained ampicillin (100 μg/ml) or erythromycin (250 μg/ml for *E. coli*, or 2.5 μg/ml for lactobacilli).", "\n\n###### \n\nList of strains, plasmids and primers used in this study.", "\n\n Strain Relevant characteristics Reference\n ---------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------- -------------------------\n *Lactobacillus paracasei* subsp. *", "paracasei* \n BGSJ2-8 BacSJ^+^, Agg^+^, EPS-SJ^+^ Laboratory collection\n BGSJ2-83 Plasmid free derivative of the strain BGSJ2-8, BacSJ^-^, Agg^-^, EPS-SJ^+^ [@B26]\n BGCG11 EPS-CG11^+^ [@B41]\n BGHN14 Agg^-^ [@B25]\n BGSJ2-83-2198-EPS7 (EPS7) Derivative of the strain BGSJ2-83, EPS-SJ^-^ This study\n *Escherichia coli* \n DH5α F^-^, Δ*lac, U169(Φ80 lacZ*Δ*M15), supE44, hsdR17, recA1, gyrA96, endA1, thi--1, relA1* [@B15]\n EC101 *supE thi* (*lacproAB*) (F9 *traD36 proAB lacI*q *Z*DM15), *repA* from pWV01 [@B28]\n ATCC25922 Serotype O6, Biotype 1 FDA strain Seattle 1946\n \n **Plasmids and constructs** **Relevant characteristics** **Reference**\n \n pA13 Em^r^, *lacZ*, plasmid pA1 derivative, 3.9 kb [@B26]\n pGhost9 Em^r^, Ts, 4.6 kb [@B35]\n pGhost9/2198 Plasmid pGhost9 derivative, carrying PCR fragment with part of the *glps_2198* gene complementary to chromosomal DNA of BGSJ2-8 This study\n pGEM-T Easy Vector 3015 bp, Amp^r^, bacterial transient vector with high level expression Promega\n \n **Primers (target)** **Primer sequence 5′--3′** **Reference**\n \n 2198 Fw (*glps_2198*) CGGGGACTGACGCCAGC This study\n 2198 Rv (*glps_2198*) GCAACCATCTCTGAAAATCCGAGG \n 2198a CAGCATCTGAGCACAGAACAG \n pGhF (pGhost plasmid) GGGGGATGTGCTGCAAGGCG This study\n pGhR (pGhost plasmid) GTCCGTTAAATCGACTGGCG \n \n\nBacSJ\n\n\\-\n\n, absence of bacteriocin BacSJ production; Agg\n\n\\-\n\n, absence of autoaggregation phenotype; EPS-SJ\n\n\\+\n\n, producer of EPS-SJ; EPS-SJ\n\n\\-\n\n, absence of EPS-SJ production; Em\n\nr\n\n, resistance to erythromycin; Amp\n\nr\n\n, resistance to ampicillin.", "\n\nSequencing and Annotation of the Putative EPS Biosynthetic Cluster\n------------------------------------------------------------------\n\nGenome sequencing was performed by Macrogen (Seoul, Korea), a next generation sequencing service. ", "The nucleotide sequence of the gene cluster *glps_2198-2211* (*epsSJ*) was deposited in the \"European Nucleotide Archive\" database^[1](#fn01){ref-type=\"fn\"}^ (Accession No. ", "LN879393). ", "An in-depth analysis of this gene cluster was performed. ", "The nucleotide sequence of each gene was used for BLAST searches against the NCBI nucleotide database. ", "When the best BLASTN hit had less than 50% coverage, a BLASTP analysis was performed.", "\n\nDNA Manipulation and Construction of EPS^-^ Mutants\n---------------------------------------------------\n\nChromosomal DNA from BGSJ2-8 was isolated as described by [@B19]. ", "For cloning, part of the *glps_2198* gene was amplified by PCR using primers 2198 Fw and 2198 Rv complementary to the *glps_2198* gene of BGSJ2-8 (**Table [1](#T1){ref-type=\"table\"}**). ", "PCR amplification was performed using a KAPA *Taq* DNA polymerase kit (Kapa Biosystems, Wilmington, MA, USA, 5 U/ml) and amplified in Gene Amp PCR system 2700 (Applied Biosystems, Waltham, MA, USA) using the following program: predenaturation (7 min, 94°C), 30 successive cycles consisting of denaturation (30 s, 94°C), annealing (1 min, 50°C), elongation (30 s, 72°C), and a final elongation (7 min, 72°C). ", "The resulting PCR amplicon, 2198 Fw/2198 Rv, was purified using QIAquick PCR Purification KIT/250 (QIAGEN GmbH, Hilden, Germany), sequenced \\[Macrogen, Europe (Amsterdam, The Netherlands)\\], and ligated into pGEM-T Easy cloning vector (Promega, Fitchburg, WI, USA). ", "General procedures for cloning and DNA manipulation were performed as described by [@B45]. ", "The ligation mixtures were used to transform *E. coli* DH5α by a standard heat-shock transformation procedure ([@B45]). ", "Transformants were selected on LB plates containing ampicillin (100 μg/ml). ", "The resulting pGEM-T Easy/2198 was sequenced. ", "Using the BLAST algorithm^[2](#fn02){ref-type=\"fn\"}^ for nucleotide sequence analysis, it was confirmed that fragment 2198 cloned into pGEM-T Easy/2198 was identical (100%) to the corresponding sequences in the genome of BGSJ2-8. ", "The construct pGEM-T Easy/2198 was further digested by *Eco*RI (Fermentas, Vilnius, Lithuania). ", "The obtained 2198 fragment (518 bp) was eluted using a QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany), and ligated into pGhost9 plasmid digested by *Eco*RI restriction enzyme. ", "The ligation mixture was used to transform *E. coli* EC101. ", "Transformants were selected on LB plates containing erythromycin (250 μg/ml). ", "The resulting construct, pGhost9/2198, carrying the appropriate fragment, was used for transformation of strain BGSJ2-83 by electroporation ([@B18]) using an Eppendorf electroporator (Eppendorf, Hamburg, Germany). ", "The BGSJ2-83/pGhost9/2198 transformants were selected on MRS plates containing erythromycin (5 μg/ml) at 28°C, and the presence of plasmid pGhost9/2198 was confirmed.", "\n\nAll plasmid isolations from *E. coli* were performed using the QIAprep Spin Miniprep kit according to the manufacturer's instructions (Qiagen, Hilden, Germany). ", "Plasmids from *L. paracasei* were isolated using the same kit with the previously described modification for *L. lactis* ([@B11]). ", "Transposition was performed as described by [@B35]. ", "A logarithmic culture of BGSJ2-83/pGhost9/2198 was grown at 28°C (a permissive temperature for the replication of plasmid pGhost9), and then transferred to 37°C for 48 h, to allow the integration of the plasmid into the host chromosome. ", "The potential EPS^-^ mutants were screened on MRS plates containing erythromycin (5 μg/ml) as erythromycin-resistant colonies, after 48 h at 37°C. ", "The integration of plasmid pGhost9/2198 into the *glps_2198* gene was confirmed by PCR using primers complementary to pGhost9 (pGhF and pGhR) and the 5′ of the *glps_2198* gene, upstream of the sequence used for the insertion mutagenesis (**Table [1](#T1){ref-type=\"table\"}**). ", "PCR was performed essentially as described above with an annealing temperature of 55°C. ", "Finally, the integration of plasmid pGhost9/2198 into the mutant's chromosome was specifically analyzed by Southern hybridization of the chromosomal DNA of strain BGSJ2-83 and its EPS^-^ derivative BGSJ2-83-2198-EPS7 (in further text EPS7) digested by *Eco*RI or *Hin*dIII by using PCR fragment 2198 as a probe. ", "Southern blot hybridization and detection of positive signals was done on the membrane, as recommended by the manufacturer (Roche Diagnostic GmbH, Mannheim, Germany).", "\n\nEPS-SJ Extraction and Purification\n----------------------------------\n\nEPS-SJ extraction and purification from strains BGSJ2-83 (*epsSJ*^+^), EPS7 (the *epsSJ* knock-out) and BGHN14 (*epsSJ*^-^) was performed essentially as described by [@B41]. ", "Bacterial strains were grown at 25°C for 48 h in a basal minimum medium ([@B3]) inoculated with an overnight culture grown in basal minimum medium (10%, v/v) and incubated. ", "After incubation, the bacteria were removed by centrifugation (12,000 × *g*, 30 min, 4°C) and EPS was extracted and precipitated at 4°C for 48 h by adding two volumes of chilled absolute ethanol. ", "The precipitate was collected by centrifugation (12,000 × *g*, 20 min, 4°C), dissolved in distilled water and dialyzed against water, using 12--14 kDa MWCO membranes (Sigma Chemical Co., St. Louis, MO, USA), for 24 h at 4°C. ", "The dialyzed retentate was freeze-dried to obtain the EPS-crude fraction which was additionally purified to reduce the DNA and protein content as previously described ([@B41]). ", "The resulting lyophilized powder was the purified EPS, which was used in different concentrations for further analyses. ", "The protein content of this purified EPS was quantified by the BCA method (Pierce, Rockford, IL, USA), the presence of DNA was determined using the a Take3 Multi-volume Plate (Biotek Instruments GmbH, Bad Friedchshall, Germany) and the EPS content was determined by means of SEC coupled with MALLS analysis (SEC-MALLS).", "\n\nSize Exclusion Chromatography\n-----------------------------\n\nThe molar mass distribution of the purified EPS was analyzed by means of a SEC-MALLS detector (Dawn Heleos II; Wyatt Europe GmbH, Dembach, Germany) as previously described ([@B41]). ", "Briefly, the chromatographic system used (Waters, Milford, MA, USA) has a photodiode array detector, set at 280 nm for detecting proteins, and a 410 refractive index detector for quantifying the different peaks obtained after separation (40°C, 0.45 ml/min, 0.1 M NaNO~3~) in the TSKgel G3000PWXL and TSKgel G5000 PWXL columns, placed in series (Supelco--Sigma, Bellefonte, PA, USA). ", "Quantification of the EPS-fractions was made from the refractive index signal using the regression equations obtained from Dextran standards of different molar mass. ", "The molar mass distribution of the EPS-fractions was obtained using Astra version 3.5 software (Wyatt Europe GmbH, Dembach, Germany).", "\n\nTransmission Electron Microscopy\n--------------------------------\n\nThe samples of bacterial colonies were encased in Bacto-agar, and the solidified blocks were cut into 1 mm^3^ cubes and fixed with 2.5% glutaraldehyde (v/v) in 0.1 M Sørensen PBS. ", "After postfixation with 2% osmium teroxide in the same buffer, samples were dehydrated using increasing concentrations of ethanol and embedded in Araldite resin (Fluka, Neu-Ulm, Germany). ", "Ultra-thin sections were obtained using a Leica UC6 ultramicrotome (Leica Microsystems, Mannheim, Germany) and mounted on copper grids. ", "For the ultrastructural analysis of the cell wall's structural components, serial sections were counterstained with 5% phosphotungstic acid in distilled water for 30 min at 56°C and subsequently with 0.1% Alcian blue in 3% acetic acid for 30 min at room temperature. ", "Sections were examined on a Philips CM12 transmission electron microscope (Philips/FEI, Eindhoven, The Netherlands) equipped with a SIS MegaView III digital camera (Olympus Soft Imaging Solutions, Hamburg, Germany). ", "The digital images were used for both cell wall ultrastructural analysis and the cell wall thickness measurements. ", "The cell wall thickness was measured by Image J software (NIH, Bethesda, MD, USA) at a magnification of 53,000×. About 50 randomly selected sections of whole bacterial cells were used for the measurements at five different regions of each cell.", "\n\nAdhesion to Caco-2 and HT29-MTX Cells\n-------------------------------------\n\nThe colonocyte-like cell lines Caco-2 and HT29-MTX were used to determine the adhesion ability of the lactobacillus strains to mucosal surfaces in the intestine. ", "A Caco-2 cell line was purchased from the European Collection of Cell Cultures (ECACC No. ", "86010202) and HT29-MTX was kindly supplied by Dr. T. Lesuffleur (INSERM UMR S 938, Paris, France; [@B30]). ", "The culture and maintenance of the cell lines were carried out as previously described ([@B41]) using Advanced DMEM medium (Gibco Invitrogen, Paisley, UK) for Caco-2 and HT29-MTX supplemented with heat inactivated fetal bovine serum (5% for Caco-2, 10% for HT29-MTX), [L]{.smallcaps}-glutamine (2 mM) and with a mixture of antibiotics (10 U/ml penicillin, 10 μg/ml streptomycin, 50 μg/ml gentamicin). ", "Media and reagents were purchased from PAA (Pasching, Austria). ", "Intestinal cells (passage 5--6) were seeded in 24-well plates and cultivated until a confluent differentiated state was reached (monolayer). ", "Lactobacillus strains were cultured for 24 h and, after washing twice with Dulbecco's PBS solution (Sigma, St. Louis, MO, USA), were resuspended in the corresponding cell-line media without antibiotics at a concentration of about 10^8^ CFU/ml. ", "The cellular monolayers were also carefully washed and lactobacillus suspensions were added at a ratio of about 10:1 (lactobacilli:eukaryotic cells). ", "The eukaryotic cells were counted in Tripan blue solution. ", "Adhesion experiments were carried out for 1 h at 37°C, 5% CO~2~ and afterward the wells were gently washed to release unattached bacteria before proceeding with the lysis of cellular monolayers using 0.25% Trypsin--EDTA solution (PAA, Pasching, Austria). ", "Dilutions of samples, before and after adhesion, were made in PBS solution and lactobacilli counts were performed in MRS agar plates. ", "The adhesion was calculated as a percentage: CFU adhered lactobacilli strain/CFU added lactobacilli strain. ", "Experiments were carried out in two replicated plates and in each plate two wells were used per sample.", "\n\nDecrease on *E. coli* Association to Caco-2 by Lactobacilli\n-----------------------------------------------------------\n\nThe capability of reference strain *E. coli* ATCC25922 to be associated to the intestinal epithelium in the presence and absence of lactobacilli was tested. ", "Bacterial cultures were washed twice with PBS and resuspended in DMEM without antibiotics at a concentration of ∼1 × 10^8^ CFU/ml; this number was corroborated by plate counting in the agar medium specific for each bacterium. ", "The bacterial suspensions containing *E. coli* ATCC25922 or a combination of *E. coli* and lactobacilli (ratio 1:1) were independently added to the Caco-2 monolayers at a ratio of 10:1, (bacteria:eukaryotic cells; in the case of *E. coli*--lactobacilli combinations, each bacterial type was added at a ratio of 5:1) and incubated at 37°C, with 5% CO~2~ for 1 h. Afterward, the monolayers were gently washed twice with PBS to remove the unattached bacteria, and the eukaryotic cells were released using 0.25% Trypsin--EDTA solution (Sigma, St. Louis, MO, USA). ", "The samples were diluted in 0.9% (w/v) NaCl buffer and plated to LA to enumerate the associated *E. coli*. ", "The percentage of *E. coli* association was calculated as follows: 100 × CFU/ml bacteria associated CFU/ml bacteria added (the dilution of bacteria was taken into account). ", "Each combination was tested in triplicate. ", "To determine the capability of the lactobacilli to decrease the association of *E. coli* ATCC25922 to Caco-2 monolayers, data were referred to that obtained with the *E. coli* alone (i.e., 100% association). ", "Each combination of *E. coli*--lactobacilli, and the *E. coli* alone as reference, were tested in three replicates.", "\n\nProliferation of GALT-Isolated Lymphocytes in the Presence of Non-Viable Lactobacilli\n-------------------------------------------------------------------------------------\n\nThe lactobacilli cells were inactivated by UV light for three cycles of 30 min each. ", "Plate counting was carried out after UV treatment to corroborate the absence of live lactobacilli that might be able to recover in the proper medium. ", "UV-inactivated lactobacilli were then divided in single use aliquots, frozen in liquid N~2~ and stored at --80°C until use ([@B32]). ", "This study was approved by the Animal Experimentation Ethical Committee of the Faculty of Pharmacy, University of Belgrade (Serbia), in strict adherence to international directives. ", "A total number of three Wistar rats (healthy female adults between 6 and 8 weeks old) were purchased from the Farm of the Military Medical Academy, Belgrade (Serbia). ", "For the experiments each animal was anesthetized with CO~2~ and, once assured of the loss of corneal reflex, its intestine was excised from the jejunum to the ileocaecal junction. ", "The whole small intestine was placed in cold Hank's balanced salt solution (HBSS without calcium and magnesium ions, prepared according to the formulation of Gibco, Invitrogen, Carlsbad, CA, USA) and kept at 4°C until processing. ", "Finally, the animals were sacrificed using the increase of CO~2~ concentration. ", "The isolation of lymphocytes from GALT (Peyer's Patches lymphocytes and intestinal epithelium lymphocytes) was carried out as previously described ([@B16]).", "\n\nTo quantify the response of GALT to the different factors tested, 2 × 10^5^ lymphocyte cells were incubated with UV-inactivated lactobacilli (at a ratio of 1:5) for 4 days in complete RPMI medium with antibiotics at 37°C and 5% CO~2~. All cultures were performed in triplicate in 96-well round-bottom microtiter plates. ", "After 4 days of incubation, the proliferation of GALT-lymphocytes was determined with a Cell Proliferation Assay Kit (Millipore Corporation, Billerica, MA, USA) following the manufacturer's instructions. ", "Results were compared to a negative control (lymphocytes growing in complete RPMI medium with antibiotics).", "\n\nStatistical Analysis\n--------------------\n\nAfter checking the normal distribution of the proliferation data, one-way ANOVA tests were used to determine the differences between each factor and the negative control. ", "Finally, one-way ANOVA tests together with the mean comparison test of least significant difference were used to compare the differences between the three strains. ", "The results are represented as mean ± standard deviation or standard error. ", "The SPSS 15.0 statistical software package (SPSS Inc, Chicago, IL, USA) was used for all determinations and *p* \\< 0.05 was considered significant. ", "The Kruskal--Wallis test (Non-parametric ANOVA) followed by Dunn's multiple comparison test were used to compare the mean values obtained at transmission electron microscopy. ", "Significant differences were set at α = 0.05.", "\n\nResults\n=======\n\nIdentification of a *glps_2198-2211* (*epsSJ*) Gene Cluster\n-----------------------------------------------------------\n\nThe entire genome of *L. paracasei* subsp. *", "paracasei* BGSJ2-8 was sequenced, annotated and partly assembled (unpublished, in house data). ", "Although BGSJ2-8 does not produce visible ropy EPS, genome annotation revealed the presence of a unique chromosomally located gene cluster *glps_2198-2211* (*epsSJ*) encoding for strain-specific EPS-SJ synthesis (**Table [2](#T2){ref-type=\"table\"}**). ", "Most of the 14 genes in the cluster showed similarities to various known proteins involved in the biosynthesis of extracellular or CPS from other lactic acid bacteria, including genes for enzymes responsible for regulation, polymerization, export, assembling repeating units, and chain length determination ([@B17]). ", "The *epsSJ* operon starts with the genes *glps_2211*, *glps_2210*, and *glps_2209* that could be putatively involved in spatial and temporal regulation of EPS-SJ synthesis and in chain length determination. ", "The *glps_2198* gene, encoding a priming glycosyltransferase, is positioned at the end of the *epsSJ* gene cluster. ", "The priming glycosyltransferase has a role in the first (priming) step in the EPS biosynthesis. ", "Upstream of the *glps_2198* there are four genes (*glps_2199*, *glps_2204*, *glps_2206*, and *glps_2207*) assigned as glycosyltransferases with function of transferring the sugars of the EPS subunit in an ordered way. ", "The gene *glps_2200*, as a putative galactoside acetyltransferase, might have be a factor modifying glycosyl residue, while the *glps_2202*, with homology to *CpsIbJ*, a putative β1, 3-galactosyltransferase, might have a role in the addition of galactose to the EPS-SJ polymer.", "\n\n###### \n\n*L. paracasei* BGSJ2-8 strain-specific EPS cluster.", "\n\n Gene ID Best BlastN/BlastP^∗^ hits Identities *E*-value Coverage\n ------------- --------------------------------------------------------------------------------- ----------------- --------------- ----------\n *glps_2198* *Lactobacillus casei* Shirota *cps1J* (AB470649), priming glycosyltransferase 498/672 (74%) 6.00*E* - 110 100%\n *glps_2199* EpsV protein *\\[Lactobacillus paracasei* subsp. *", "paracasei* 8700:2\\]\\# 156/315 (49%)\\# 8*e* - 85\\# 98%\\#\n *glps_2200* *Lactobacillus casei* Shirota *cps1G* (AB470649), galactoside acetyltransferase 412/574 (71%) 3.00*E* - 75 85%\n *glps_2201* *Streptococcus thermophilus* LY03 *eps* gene cluster; 131/187 (70%) 4.00*E* - 12 18%\n membrane protein \\[*Oenococcus oeni* ATCC BAA-1163\\]\\# *137/314\\#* *1e* - *69*\\# *67%\\#*\n *glps_2202* *CpsIbJ* \\[*Streptococcus agalactiae*\\]\\# 81/229 (35%)\\# 5e - 39\\# 73%\\#\n *glps_2203* *EpsL* \\[*Streptococcus thermophilus*\\]\\# 95/315 (30%)\\# 2e - 35\\# 99%\\#\n *glps_2204* glycosyltransferase \\[*Lactobacillus paracasei* subsp. *", "paracasei* 8700:2\\]\\# 120/385 (31%)\\# 8e - 39\\# 98%\\#\n *glps_2205* predicted protein \\[*Lactobacillus paracasei* subsp. *", "paracasei* 8700:2\\]\\# 97/364 (26%)\\# 2e - 15\\# 92%\\#\n *glps_2206* *Oenococcus oeni* PSU-1, glycosyltransferase 255/395 (64%) 2.00*E* - 11 46%\n *glps_2207* *Leuconostoc citreum* KM20, glycosyltransferase 670/992 (67%) 2*e* - 81 81%\n *glps_2208* *Plasmodium falciparum* 3D7 23/23 (100%) 0.58 13%\n *glps_2209* *Lactobacillus rhamnosus* RW-6541M, Wze (AY659978); 558/741 (75%) 3.00*E* - 139 99%\n *Lactobacillus casei* Shirota *cps1B* (AB470649), chain length determination 523/690 (75%) 2.00*E* - 135 92%\n *glps_2210* *Lactobacillus rhamnosus* RW-6541M, *wzd* (AY659978) 265/382 (69%) 5.00*E* - 33 81%\n *glps_2211* *Lactobacillus rhamnosus* GG, chain length regulator 258/349 (73%) 1.00*E* - 52 84%\n *Lactobacillus casei* Shirota *cps1A* (AB470649), chain length determination 247/345 (71%) 3.00*E* - 41 84%\n \n\n∗\n\nWhen the best BLASTN hit has less than 50% coverage, a BLASTP analysis was performed. ", "Best BlastP hit is shown in the table; \\#The scores of BlastP results.", "\n\nInterestingly, several genes in the *epsSJ* gene cluster have the best BLAST hits to the gene cluster for biosynthesis of the CPS previously described in probiotic *L. casei* Shirota (CPS operon Shirota; [@B54]) (**Figure [1](#F1){ref-type=\"fig\"}**). ", "The nucleotide sequences of genes *glps_2198* (518 bp), *glps_2200* (435 bp), *glps_2209* (524 bp), and *glps_2211* (269 bp) of strain BGSJ2-8 showed identity to the nucleotide sequences of the genes *cps1J* (74%), encoding primary glycosyltransferases, *cps1G* (71%), encoding galactoside acetyltransferase, and *cps1B* (75%) and *cps1A* (71%), encoding polysaccharide chain length determination proteins, within the Shirota CPS operon.", "\n\n![**", "Schematic representation of CPS and EPS-SJ synthesis clusters in *L. casei* Shirota and *L. paracasei* BGSJ2-8, respectively.** ", "Genes sharing high sequence similarity are indicated and highlighted in the same colors. ", "Adapted from [@B54].](fmicb-07-00286-g001){#F1}\n\nConstruction and Analysis of the EPS-SJ^-^ Mutants\n--------------------------------------------------\n\nIn order to determine the functional properties of the annotated *epsSJ* gene cluster, the gene encoding the primary glycosyltransferase *glps_2198* was disrupted by insertional inactivation as described in Section \"Materials and Methods.\" ", "Initially, the isolation of plasmid DNA from the erythromycin resistant colonies obtained after the transposition experiment indicated that the construct pGhost9/2198 had been integrated into the chromosome of BGSJ2-83. ", "The experimental work was continued with a strain designated as EPS7, a potential mutant in the *glps_2198* gene encoding the primary glycosyltransferase. ", "EPS7 showed a similar growth phenotype to the parental strain (data not shown). ", "The chromosomal integration of plasmid pGhost9/2198 was initially verified by isolation of plasmid DNA from the EPS7 mutant and transformation of competent *E. coli* EC101. ", "After 24 h of incubation at 37°C, transformants were not obtained, indicating the absence of circular forms of plasmid pGhost9/2198 in the EPS7 mutant strain. ", "Additionally, the integration was confirmed by PCR and Southern blot hybridization (as described in section \"Materials and Methods\"). ", "The results revealed that plasmid pGhost9/2198 had been integrated into the chromosomal DNA of EPS7 at the specific site of the *glps_2198* gene (data not shown).", "\n\nMolar Mass Distribution of EPS\n------------------------------\n\nSEC-MALLS analysis was performed to analyze the molar mass distribution of the purified polymeric material and to quantify the EPS fractions. ", "The strain *L. paracasei* subsp. *", "paracasei* BGHN14 was used as a negative reference strain, since PCR analysis confirmed that the *epsSJ* genes are not present in this strain (data not shown).", "\n\nThe results showed that the polymer purified from strain BGSJ2-83 presents two EPS fractions synthesized in similar amounts (**Figure [2](#F2){ref-type=\"fig\"}**). ", "One has a higher molecular weight (P1), about 41,200 Da, whereas the second fraction is smaller, about 13,600 Da (P2). ", "However, it is worth mentioning that the P2 fraction might also contain proteins, since in the elution time corresponding to this peak (around 38 min) there was also an increase in the signal corresponding to the photodiode array detector in the UV spectrum (280 nm). ", "Interestingly, in the EPS purified from EPS7 (the mutant), in which the primary glycosyltransferase *glps_2198* gene was disrupted by insertional inactivation, the P2 fraction was not detected (although the increase in the UV signal is present), while the P1 fraction is present in lower amount (2/3 of amount detected in BGSJ2-83) and with higher molecular weight (about 80,000 Da). ", "In the case of EPS purified from strain BGHN14, lacking the entire *epsSJ* gene cluster, the highest production corresponded to the fraction coinciding with the P2 EPS-SJ (1.6 higher amount than in BGSJ2-83), also presenting the same increase in the UV signal. ", "Indeed, the amount of protein quantified by BCA was similar in both EPS fractions, i.e., 15.6 ± 0.5 μg/ml and 15.1 ± 0.3 μg/ml for BGSJ2-83 and BGHN14, respectively.", "\n\n{#F2}\n\nElectron-Microscopic Analysis and Cell Wall Thickness Measurement\n-----------------------------------------------------------------\n\nBoth, electron-microscopic analysis (**Figure [3A](#F3){ref-type=\"fig\"}**), and cell wall thickness measurement (**Figure [3B](#F3){ref-type=\"fig\"}**) of *L. paracasei* subsp. *", "paracasei* strains, showed a significant decrease of cell wall thickness in EPS7 and BGHN14 strains compared to BGSJ2-83 (∼40 nm). ", "Indeed, it was observed that EPS7, lacking the P2 EPS fraction (∼30 nm), had even thinner cell wall than BGHN14 lacking P1 (∼35 nm).", "\n\n{#F3}\n\nIn addition, ultrastructural analysis of the different strains of *L. paracasei* subsp. *", "paracasei* demonstrated an underlying alteration in cell wall structure. ", "In BGHN14 and EPS7, the cell walls appeared pale, with low electron-density, particularly in the EPS7 mutant. ", "The outer surface of the cell wall of these groups was smooth, without visible protrusions. ", "Quite the opposite, the cell wall of BGSJ2-83 had a dark, electron-dense appearance. ", "Also, the cell wall of BGSJ2-83 was slightly curved in shape with additional protrusions of fibrillar polysaccharide material at the extracellular surface.", "\n\nFurthermore, detailed examination revealed the presence of a thin, electron-dense layer at the outermost surface of the cell wall in all of the groups, suggesting the presence of proteins, particularly in EPS7 and BGHN14, where this layer was clearly visible. ", "In contrast, the presence of proteins in BGSJ2-83 was harder to resolve, and in most cases it was visibly absent.", "\n\nCharacterization of the Role of EPS-SJ in Adhesion to Epithelial Intestinal Cells\n---------------------------------------------------------------------------------\n\nTo estimate the role of EPS-SJ in the adhesion ability of strain BGSJ2-8 to IEC, *in vitro* comparative analysis of strain BGSJ2-83 and its EPS-SJ^-^ derivative EPS7 was performed. ", "The strains BGCG11, producing high molecular weight EPS-CG11, and BGHN14, lacking *epsSJ* gene cluster, were used as controls. ", "In particular, the EPS-SJ-producing strain BGSJ2-83 and strain BGHN14 showed higher percentages of adhesion to the Caco-2 cell line than the mutant, EPS7 (**Figure [4A](#F4){ref-type=\"fig\"}**).", "\n\n{#F4}\n\nIn parallel, the adhesion ability of strains BGSJ2-83 and EPS7 was tested on the HT29-MTX cell line, characterized by mucus production, which is the first barrier to the interaction of bacteria with IEC ([@B30]). ", "Again, BGSJ2-83 showed a higher percentage of adhesion to HT29-MTX cells than EPS7, although in this case BGHN14 showed low adhesion ability, similar to EPS7 (**Figure [4A](#F4){ref-type=\"fig\"}**).", "\n\nCharacterization of the Role of EPS-SJ in Decrease on *E. coli* Association to Caco-2 Cells\n-------------------------------------------------------------------------------------------\n\nIn this study, we analyzed the potential of EPS-SJ-producing and non-producing strains to compete with *E. coli* and to reduce the association capability of *E. coli* to Caco-2 epithelial cells (**Figure [4B](#F4){ref-type=\"fig\"}**). ", "The *E. coli* strain ATCC25922 was significantly less associated to Caco-2 cells in the presence of BGSJ2-83 and control strain BGCG11 than when it was analyzed alone, indicating the ability of all lactobacilli strains to diminish the association of *E. coli* to Caco-2 cells. *", "E. coli* exhibited the lowest association rate to the Caco-2 cell line when co-incubated with the strain BGSJ2-83, which produces EPS-SJ, as well as in the presence of control strain BGCG11. ", "No significant differences were detected between the two strains. ", "In contrast, when *E. coli* was co-incubated with EPS7 the association of *E. coli* to Caco-2 cells was reduced, but not to the same extent as in the presence of the EPS producing strains.", "\n\nEffect of Insertional Mutagenesis on the Role of EPS-SJ on Proliferation of GALT-Isolated Lymphocytes\n-----------------------------------------------------------------------------------------------------\n\nIn order to monitor the effect of EPS-SJ's presence on the cell surface of BGSJ2-83, the proliferation of the primary GALT culture in the presence of UV-irradiated cells of BGSJ2-83 and EPS7 was followed during 3 days of co-incubation. ", "The strains BGCG11, producing high molecular weight EPS-CG11, and BGHN14, lacking *epsSJ*, were used as reference strains. ", "The results are expressed as a percentage relative to the cells incubated in RPMI medium alone, with a standard error (**Figure [4C](#F4){ref-type=\"fig\"}**). ", "The results show that UV-irradiated cells of both BGSJ2-83 and EPS7 induced lower levels of lymphocyte proliferation compared to the control GALT-isolated lymphocytes. ", "Interestingly, the level of proliferation of GALT-isolated lymphocytes determined for the EPS7 mutant was higher compared to BGSJ2-83, indicating that EPS-SJ P1, present on the cell surface of BGSJ2-83 and absent on the cell surface of EPS7, contributed to the lower proliferation of GALT-isolated lymphocytes.", "\n\nDiscussion\n==========\n\nIn this study, we have investigated the role of exopolysaccharide EPS-SJ, produced by the natural dairy isolate *L. paracasei* subsp. *", "paracasei* BGSJ2-8, in adhesion to IEC, decreased association of *E. coli* to Caco-2 cells and proliferation of GALT-isolated lymphocytes. ", "Our results show that the unique *epsSJ* gene cluster identified in the chromosome of *L. paracasei* subsp. *", "paracasei* BGSJ2-8 is involved in the biosynthesis of strain-specific EPS-SJ, containing two EPS fractions, P1 and P2. [", "@B54] remarked that *L. casei* Shirota produced EPS with two peaks of different molecular weight, named polysaccharide-peptidoglycan (PS-PG) polymer. ", "In addition, [@B54] showed that the mutation of different genes in the cluster from *cps1A* to *cps1J* modified the synthesis of high molecular weight-EPS (named PS1), but a smaller fraction (PS2) remained intact.", "\n\nIn our study, the results of a SEC-MALLS analysis revealed that insertional mutagenesis of the *glps_2198* gene, encoding priming glycosyltransferase, led to the absence of the P2 fraction in EPS7, related to low molecular weight EPS-SJ (13,600 Da) present in BGSJ2-83, as well as to an increase in molecular weight of the P1 and its reduced production. ", "According to literature data the gene encoding priming glycosyltransferase is highly evolutionarily conserved ([@B21]) and could be located at the beginning of the *eps* operons, as reported for *L. helveticus* NCC2745 ([@B20]) and *L. paraplantarum* BGCG11 ([@B56]), in the middle of the *eps* clusters, as in *Streptococcus thermophilus* strains ([@B51]; [@B1]), *L. lactis* NIZO B40 ([@B24]), and *L. delbrueckii* subsp. *", "bulgaricus* Lfi5 ([@B27]), or in the vicinity of the end of the *eps* clusters as in the case of *L. rhamnosus* ([@B42]) and here in the case of BGSJ2-8. ", "The priming glycosyltransferase is involved in the first step of biosynthesis of an oligosaccharide \"block.\" ", "It is a member of the EPS biosynthesis polyprenyl glycosyl phosphotransferases family with the role of transferring the sugar from UDP-glucose or UDP-galactose to a lipid carrier. ", "The results obtained in this study revealed that interruption of the *glps_2198* gene completely affected the biosynthesis of the low molecular weight P2 fraction of EPS-SJ, rather than the high molecular weight P1 fraction, although the P1 fraction is synthesized in lower amount. ", "The results are in accordance to work of [@B38] where it was reported that low molecular weight EPS are more common in lactic acid bacteria than high molecular weight. ", "Hence, it could be deduced that the gene *glps_2198* encoding priming glycosyltransferase in BGSJ2-83 is necessary for the first steps in biosynthesis of EPS-SJ.", "\n\nIncreased electron-density as a consequence of specific polysaccharide phosphotungstic acid/Alcian blue counterstaining suggests a significant ratio of EPS in the cell wall of this strain. ", "These fibrillar protrusions probably correspond to EPS-SJ, the presence of which was subsequently confirmed by SEC-MALLS analysis, detected in BGSJ2-83.", "\n\nThe results from electron-microscopic analysis and cell wall measurement clearly indicate that both fractions of EPS-SJ, present on the cell surface of BGSJ2-83, contribute to the thickness of the cell wall in this strain. ", "In fact, the absence of the P1 and P2 fractions in BGHN14 and EPS7, respectively, influenced the decrease of cell wall thickness by approximately 5 and 10 nm. ", "Besides its lower content, this could also be a consequence of increased electron-density of the peptidoglycan layer due to increased polysaccharide content.", "\n\nProbiotic lactobacilli exert their positive effect on human health by various mechanisms. ", "The adhesive properties of probiotic strains provide several important abilities: the colonization of intestinal mucosa, persistence in the intestine, the competitive exclusion of pathogenic microorganisms, and immunomodulation ([@B39]). ", "In fact, the colonization of intestinal mucosa is an important criterion proposed by FAO/WHO guidelines for the selection of strains with probiotic potential, since their health-promoting effects might be partly dependent on their persistence in the intestine and adhesion to mucosal surfaces ([@B4]). ", "Hence, the influence of EPS-SJ on the potential to adhere to the intestinal mucosal surface was examined using widely accepted *in vitro* models, Caco-2 and HT29-MTX cells.", "\n\nIt has been suggested that the adhesive properties of lactobacilli largely depend on the surface properties of probiotic strains (e.g., EPS), enabling them to interact with the mucosa and IEC ([@B48]). ", "It can be assumed that the better adhesion of BGSJ2-83 and BGHN14 than the EPS7 mutant to Caco-2 cells is probably due to the presence of EPS-SJ P2 on the cell surface of strains BGSJ2-83 and BGHN14, which ensures better communication between bacteria and IEC. ", "In contrast, the strain BGCG11, producing ropy EPS-CG11, was previously described to have low adhesion ability, which was correlated with the presence of EPS-CG11 on the cell surface ([@B41]).", "\n\nIt has been shown previously that the adhesion of bacteria to IEC is better in the absence of cell surface polysaccharides, due to better exposure of cell surface proteins in bacteria to the adhesin in eukaryotic cells ([@B44]; [@B7]; [@B41]; [@B43]; [@B8]). ", "Interestingly, EPS-SJ P2 is most likely positively involved in the adhesion of the BGSJ2-8 strain to IEC. ", "Specifically, our results revealed that the presence of the EPS-SJ P2 around the bacteria increases adhesion to Caco-2 cells, in contrast to previously published data. ", "Particularly, [@B41] showed that presence of EPS-CG11 negatively affected adhesion of *L. paraplantarum* BGCG11 to IEC. ", "In addition, [@B43] suggested that polysaccharides produced by *L. rhamnosus* E/N interfere with adhesion by masking protein receptors at the cell surface. ", "However, the results of [@B44] demonstrated that the adhesion of bacteria to IEC depends on the EPS type and the dose. ", "Such, the EPS GG produced by *L. rhamnosus* GG at the highest dose increased the adherence of *Bifidobacterium animalis* IPLA-R1. ", "Probably it could be referred to differences in physicochemical and/or structural characteristics of the EPS polymers, as well as to the surface characteristics of the strains that contribute to the differences in adhesion ([@B44]). ", "Hence, it is likely that the molecular structure of EPS-SJ P2 matrix contributes to its adhesion to IEC, although the role of EPS molecules in the adhesion of bacteria to IEC still remains unclear.", "\n\nDue to the fact that the presence of mucin in the HT29-MTX cell line simulates the bacterial interaction with gut mucus ([@B36]), the adhesion to HT29-MTX cells was also tested. ", "In general, the tested strains showed a higher percentage of adhesion to the HT29-MTX cell line, while their adhesion ability to the Caco-2 cell line was lower. ", "Similarly, the adhesion levels of strains *L. rhamnosus* DR20, *L. acidophilus* HN017, and *Bifidobacterium lactis* DR10 were two to three times higher to HT29-MTX cells compared to the degree of adhesion to Caco-2 cells ([@B14]). ", "These findings are in contrast to our previous results obtained with other putative probiotic strains, where the adhesion to Caco-2 cells was significantly higher than to HT29-MTX cells ([@B52]). ", "It is possible that the presence of the glycoprotein layer (mucin) on the surface of HT29-MTX cells participates in the interaction with the cell surface compounds of bacterial cells, which contribute to their hydrophobicity. ", "Furthermore, the adhesion of strain BGHN14 and the EPS7 mutant, lacking the P1 and P2, respectively, was lower than the adhesion of BGSJ2-83 to HT29-MTX cells. ", "However, there is no significant difference between the adhesion of BGHN14 and EPS7 strains to HT29-MTX, indicating that EPS-SJ P2 more specifically contributes to the adhesion of BGSJ2-83 and EPS7 to IEC than to the gut mucus matrix.", "\n\nThe other important health promoting property of probiotic strains is their capability to neutralize the negative effects of pathogens ([@B10]). ", "This characteristic is shown to be strain-dependent and several mechanisms of action have been proposed to explain it: (i) a physical blocking of the pathogen's entry (colonization competition), (ii) induction of mucus production, (iii) reinforcement of the selective permeability of the epithelium by increasing tight-junctions, (iv) production of antimicrobial factors, and, among others, (v) stimulation of the innate immune response ([@B13]; [@B31]). ", "In this study we have shown that EPS-SJ-producing strain BGSJ2-83 exhibits higher adhesion to Caco-2 cells and is more capable of competing for the colonization of the intestinal niche than *E. coli* strain ATCC25922. ", "Similarly, non-EPS producing derivative of EPS-CG11 (NB1) more efficiently reduced the association of *Clostridium difficile* or *Yersinia enterocolitica* than parental BGCG11 strain ([@B56]). ", "Interestingly, the presence of EPS-CG11-producing strain BGCG11, exhibiting low adhesion to Caco-2 cells, decreased the association of *E. coli* ATCC25922 to Caco-2 cells, similarly to previously obtained results by [@B56]. ", "Hence, it could be concluded that both EPS-CG11 and EPS-SJ are effective macromolecules protecting the IEC from pathogen binding regardless of their adhesion ability ([@B56]). ", "It is most likely that *E. coli* cells failed to access the IEC surface because they were trapped within the EPS matrix. ", "It has been previously reported that different lactobacillus species have the ability to reduce the association of diverse enteropathogens to intestinal cells ([@B12]). ", "This capability has also been attributed to antimicrobial activity related to the production of bacteriocins or organic acids ([@B46]; [@B23]). ", "Since strain BGSJ2-83 is a BGSJ2-8 derivative lacking the plasmid carrying the genes responsible for bacteriocin production and does not exhibit antimicrobial activity, it is more likely that its capability is related to the presence of specific structural components on the cell surface. ", "It has been proposed that the cell surface components involved in this feature could be cell surface associated proteins ([@B53]), S-layer macromolecules built from proteins ([@B55]) and aggregation factors ([@B23]), as well as EPS ([@B56]). ", "Hence, the results obtained in this study indicate that EPS-SJ could be involved in the decreased association of *E. coli* to Caco-2 cells. ", "Since the results of this study show that association of *E. coli* with Caco-2 cells is higher in the presence of EPS7 than in the presence of BGSJ2-83, it can be concluded that EPS-SJ P2, lacking in EPS7, specifically contributes to this feature.", "\n\nFinally, an important feature of potential probiotic candidates is the capacity to modulate the host's immune response either in the terms of strengthening the immunological barrier against pathogen, or decreasing the immune responsiveness in exaggerated inflammatory conditions. ", "The study on *L. casei* Shirota's *cps* gene cluster has shown that this operon determines the synthesis of a specific polysaccharide moiety relevant for immune regulation ([@B54]). ", "The production of high molecular weight PS1 was directly related to the suppressive effect of this strain upon IL-6 production in macrophages stimulated with LPS ([@B54]). ", "The results of our study indicate that the presence of EPS-SJ on the cell surface contributes to the lower level of GALT-isolated lymphocyte proliferation. ", "The level of GALT-isolated lymphocyte proliferation stimulated with EPS7 mutant was higher compared to BGSJ2-83, indicating that EPS-SJ P2, present on the cell surface of BGSJ2-83, has an inhibitory effect on GALT-isolated lymphocyte proliferation. ", "The GALT is intestinal frontier and essential component of systemic immune response, which protects body from pathogens and foreign antigens. ", "On the other hand it enables tolerance to commensal bacteria and dietary antigens and protects mucosa from detrimental inflammatory immune responses ([@B50]; [@B22]). ", "Hence, GALT has a pivotal role in the decision between immune response and tolerance. ", "The lower level of GALT-isolated lymphocytes proliferation in the presence of EPS-SJ-producing strain, BGSJ2-83, could indicate modulation of the immune response toward the tolerogenic environment. ", "This feature could be useful in treatment of excessive inflammatory condition such as Crohn's disease and other inflammatory disorders which occur as a result of an inappropriate innate and/or adaptive immune response. ", "However, further experiments are needed in order to decipher the molecular mechanism of this feature.", "\n\nConclusion\n==========\n\nThe results obtained in this study indicate that EPS-SJ, specifically the P2 fraction, interacts with a certain host cell component, providing better gut colonization and pathogen exclusion. ", "These results may contribute to the previous knowledge of the role of EPS in adhesion to IEC. ", "Namely, it was established previously that polysaccharides hindered bacterial adhesion to host by masking protein receptors involved in the adherence to IEC. ", "Here, we report that EPS may positively or negatively affect adherence of bacteria to the host cells.", "\n\nAuthor Contributions\n====================\n\nMŽ -- conception and design of the work, performed main work, analyzed, interpreted and critically revised the data; MSM and MT -- construction and characterization of the knock-out, draft of the work related to molecular genetics; PR-M -- conception and design of SEC-MALLS, performed the SEC-MALLS analyses, draft related to SEC-MALLS and critically revised the manuscript; MBM -- performed work and made part of the draft related to transmission electron microscopy; KV and MT -- performed a work related to adhesion, *E. coli* association to Caco-2 cells, and GALT proliferation, made part of the draft and critically revised the manuscript; SS -- EPS characterization; AK -- conception and design, performed the work related to transmission electron microscopy, made the draft and critically revised the manuscript; NG -- conception and the design of the work, supervised the complete work, analyzed and interpreted the data, draft the work, critically revised the manuscript. ", "All authors finally approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.", "\n\nConflict of Interest Statement\n==============================\n\nThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.", "\n\n**Funding.** ", "This work was funded by the Ministry of Education, Science and Technological Development of the Republic of Serbia, grant OI 173019. ", "The founders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.", "\n\nT. Lesuffleur (INSERM U843, Paris, France) is acknowledged for the kind supply of HT29-MTX cellular line. ", "Mengjin Liu and Roland Siezen are acknowledged for the *Lactobacillus paracasei* subsp. *", "paracasei* BGSJ2-8 genome annotation. ", "We are grateful to Nathaniel Aaron Sprinkle, native English editor for the proofreading of the manuscript.", "\n\n<http://www.ebi.ac.uk/ena>\n\n<http://www.ncbi.nlm.nih.gov/BLAST>\n\nCFU\n\n: colony forming units\n\nCPS\n\n: capsular polysaccharides\n\nEPS\n\n: extracellular polysaccharides\n\nIEC\n\n: intestinal epithelial cells\n\nGALT\n\n: gastrointestinal associated lymphoid tissue\n\nLB\n\n: Luria Broth\n\nMALLS\n\n: multi-angle laser light scattering\n\nMRS\n\n: De Man-Rogosa-Sharpe medium\n\nPBS\n\n: phosphate buffer solution\n\nSEC\n\n: size exclusion chromatography\n\n[^1]: Edited by: *Andrea Gomez-Zavaglia, Center for Research and Development in Food Cryotechnology -- National Scientific and Technical Research Council, Argentina*\n\n[^2]: Reviewed by: *Michael Gänzle, University of Alberta, Canada; Micaela Medrano, Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina*\n\n[^3]: This article was submitted to Food Microbiology, a section of the journal Frontiers in Microbiology\n" ]

{ "pile_set_name": "PubMed Central" }

[ 0.005917159763313609, 0, 0, 0.006493506493506494, 0.009259259259259259, 0.004016064257028112, 0.005208333333333333, 0.006535947712418301, 0.015503875968992248, 0.015384615384615385, 0.007194244604316547, 0.015748031496062992, 0.013245033112582781, 0.011627906976744186, 0.0053475935828877, 0.009615384615384616, 0.004273504273504274, 0.0070921985815602835, 0.0049261083743842365, 0.0040650406504065045, 0.02027027027027027, 0.02564102564102564, 0.010582010582010581, 0.030303030303030304, 0.011904761904761904, 0.0055248618784530384, 0.015625, 0.015810276679841896, 0.013888888888888888, 0, 0.012195121951219513, 0.013986013986013986, 0.027522935779816515, 0.007194244604316547, 0.0048543689320388345, 0.013245033112582781, 0.005128205128205128, 0, 0, 0, 0.005141388174807198, 0.00425531914893617, 0, 0, 0, 0.009708737864077669, 0.011764705882352941, 0.011560693641618497, 0.005376344086021506, 0.00980392156862745, 0.015037593984962405, 0.01098901098901099, 0.008333333333333333, 0.013157894736842105, 0, 0.004347826086956522, 0.020833333333333332, 0.00546448087431694, 0, 0.01282051282051282, 0.004672897196261682, 0.006024096385542169, 0.006134969325153374, 0.007633587786259542, 0.019230769230769232, 0, 0.006802721088435374, 0.0035971223021582736, 0.011363636363636364, 0.003205128205128205, 0.012048192771084338, 0.016194331983805668, 0.005780346820809248, 0.00510204081632653, 0.0044444444444444444, 0.011299435028248588, 0.008333333333333333, 0.025078369905956112, 0.0163265306122449, 0.007832898172323759, 0.006024096385542169, 0.015037593984962405, 0, 0.005319148936170213, 0.014705882352941176, 0, 0.013888888888888888, 0, 0.004098360655737705, 0.008298755186721992, 0.022222222222222223, 0.018691588785046728, 0.014962593516209476, 0.015625, 0, 0.00819672131147541, 0, 0, 0.00784313725490196, 0.014925373134328358, 0.018518518518518517, 0, 0.007142857142857143, 0.01327433628318584, 0.005357142857142857, 0.009345794392523364, 0.005780346820809248, 0, 0.004807692307692308, 0, 0.0038461538461538464, 0, 0.015037593984962405, 0.01098901098901099, 0.011976047904191617, 0.005555555555555556, 0.017391304347826087, 0.0125, 0.019230769230769232, 0.006211180124223602, 0.004901960784313725, 0.009345794392523364, 0.004629629629629629, 0.006097560975609756, 0, 0.006756756756756757, 0.017142857142857144, 0, 0.005434782608695652, 0, 0.011904761904761904, 0.006309148264984227, 0.004830917874396135, 0, 0.010416666666666666, 0.0045871559633027525, 0.0036101083032490976, 0.016129032258064516, 0.0020491803278688526, 0.0012004801920768306, 0, 0.0029006526468455403, 0, 0.011857707509881422, 0.002288329519450801, 0, 0.015625, 0, 0.007653061224489796, 0.004545454545454545, 0, 0, 0, 0, 0.007462686567164179, 0, 0.014492753623188406, 0, 0.012578616352201259, 0.012121212121212121, 0, 0, 0.0026041666666666665, 0.007662835249042145, 0.024242424242424242, 0, 0.04411764705882353, 0.0125, 0.008281573498964804, 0.015267175572519083, 0.007575757575757576, 0, 0.010309278350515464, 0.0196078431372549, 0, 0, 0, 0, 0, 0.0196078431372549, 0.01694915254237288, 0, 0.017793594306049824, 0, 0.00909090909090909, 0, 0.011764705882352941, 0, 0.003816793893129771, 0, 0.014367816091954023, 0.023622047244094488, 0.010362694300518135, 0, 0, 0.025, 0, 0.09090909090909091, 0.07692307692307693, 0, 0.125, 0.015463917525773196, 0.011363636363636364, 0, 0, 0, 0.043478260869565216, 0, 0, 0.014084507042253521, 0, 0, 0, 0.09090909090909091, 0.07692307692307693, 0, 0.09090909090909091, 0, 0.01312910284463895, 0.005076142131979695, 0.011876484560570071, 0.01079136690647482, 0.015706806282722512, 0, 0.010638297872340425, 0.006772009029345372, 0.024390243902439025, 0.006329113924050633, 0, 0.016129032258064516, 0.0125, 0.014388489208633094, 0, 0.025, 0.02, 0.004694835680751174, 0.011235955056179775, 0.01411764705882353, 0.01948051948051948, 0, 0.011111111111111112, 0.0035460992907801418, 0.011904761904761904, 0.012422360248447204, 0.005235602094240838, 0.013157894736842105, 0.008888888888888889, 0.006289308176100629, 0, 0, 0.004201680672268907, 0.006622516556291391, 0.011627906976744186, 0.014705882352941176, 0.019157088122605363, 0.020833333333333332, 0.019157088122605363, 0.009433962264150943, 0.011904761904761904, 0.025, 0.00641025641025641, 0.01680672268907563, 0.015384615384615385, 0.008583690987124463, 0.02030456852791878, 0.005555555555555556, 0.006211180124223602, 0.008658008658008658, 0.01020408163265306, 0, 0.025, 0.01282051282051282, 0.006802721088435374, 0.004395604395604396, 0.013761467889908258, 0.02072538860103627, 0.022321428571428572, 0.022727272727272728, 0.008264462809917356, 0.005917159763313609, 0.013888888888888888, 0, 0.02066115702479339, 0.014285714285714285, 0.012145748987854251, 0, 0.005494505494505495, 0.011627906976744186, 0.01282051282051282, 0.020080321285140562, 0.007042253521126761, 0.011976047904191617, 0.011627906976744186, 0.015151515151515152, 0.0045662100456621, 0, 0.004629629629629629, 0.02127659574468085, 0, 0.009900990099009901, 0.011684518013631937, 0, 0, 0, 0.007518796992481203, 0, 0.009259259259259259, 0.02247191011235955, 0, 0.009433962264150943, 0.019384264538198404 ]

[ "Q:\n\nselenium button click not triggering inline javascript event handler\n\nMy HTML looks like below\n<div class=\"container\">\n <div class=\"jumbotron\">\n <h1 id=\"survey-title\" class=\"display-5 text-capitalize\">Your favourite candidate</h1>\n <hr class=\"my-4\">\n\n <div class=\"card-body\">\n <h5 id=\"question-title\" class=\"card-title\">What is your name?</h5>\n\n <form id=\"question_form\" action=\"/survey/takesurvey/1/2\" method=\"post\" class=\"form\">\n <input type=\"hidden\" name=\"csrfmiddlewaretoken\" value=\"ELBWYGZRxRuI7CoZ2xkmgCJ3f9JweFMM4Ew1pQbgSE3BLb38VStPJEbHvsyiBEFg\">\n <div class=\"form-group\"><input type=\"text\" name=\"response\" value=\"asdfasdfas\" maxlength=\"400\" class=\"form-control\" title=\"\" id=\"id_response\"></div>\n </form>\n </div>\n\n <a id=\"btn-previous\" href=\"javascript:history.back()\" class=\"btn btn-primary \">Previous</a>\n <a id=\"btn-next\" href=\"/survey/takesurvey/finish/1\" class=\"btn btn-primary\">Next</a>\n </div>\n</div>\n\nNotice the href on previous button\n<a id=\"btn-previous\" href=\"javascript:history.back()\" class=\"btn btn-primary \">Previous</a>\nClicking next will submit form and go to \"next\" page, again clicking \"previous\" should come back and preserve the input value in previous page. ", "I am using following code to test this behavior, The following code is working only when I put sleep(1)line in between finding the element and clicking. ", "Else the page is not going back (url not changing)\n\nWhy just waiting for url change before clicking the button not working\nI suspect it has to do with javascript event handler (which is inline in html) not being loaded.", "\nI is there any way to check if page is ready for the click event\n\ncur_url = self.browser.current_url\nself.browser.find_element_by_id('btn-next').click()\n\nwait(lambda: self.assertNotEqual(self.browser.current_url, cur_url))\nbtn_previous = self.browser.find_element_by_id('btn-previous')\n\n## This code is not working without the sleep below. ", "Which I suspect \n## has to do with javascript event handler (which is inline in html) not being loaded.", "\n## Is there better way to check that the button is ready to be clicked\nsleep(1)\n\nbtn_previous.click()\n\nA:\n\nYou can use WebDriverWait with element_to_be_clickable condition. ", "Check here also.", "\nfrom selenium.webdriver.common.by import By\nfrom selenium.webdriver.support.ui import WebDriverWait\nfrom selenium.webdriver.support import expected_conditions as EC\n\nWebDriverWait(self.browser, 5).until(EC.element_to_be_clickable((By.", "ID, \"btn-previous\"))).click()\n\nYou can try wait for JavaScript:\nWebDriverWait(self.browser, 20).until(lambda d: d.execute_script(\n 'return (document.readyState == \"complete\" || document.readyState == \"interactive\")'))\n\n# or without interactive \nWebDriverWait(self.browser, 20).until(lambda d: d.execute_script(\n 'return document.readyState == \"complete\"'))\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0, 0, 0, 0, 0, 0, 0.01276595744680851, 0.005376344086021506 ]

[ "BY\nTom Parry\n\nShare\n\nTravel ‘a career for life’ despite lag in salaries\n\nMore than 60% of rising travel leaders consider industry salaries to be lagging behind those offered in other sectors. ", "But according to new research by TTG Media, 81% still plan to spend their entire working life in the industry.", "\n\nSharelines\n\n“People think it’s a party but it’s not – you have to work hard, but my god you will love it.”", "\n\nThe findings, which were drawn from the responses of around 90 members of TTG’s 30 Under 30 initiative, were revealed during WTM London as part of the TTG Tomorrow’s Travel Leaders Conference.", "\n\nIt was the first time the forum, designed to advise the next generation of senior travel professionals, was held at the trade show.", "\n\nPresenting the research, Daniel Pearce, managing director of TTG Media, and group editor Pippa Jacks revealed 70% of respondents – whose experience working in travel ranged from 1 to 13 years – would consider launching their own travel business. ", "It was also revealed that, in spite of the perceived lower pay in travel, 100% of respondents would encourage a young person beginning their working life to join the sector.", "\n\nDelegates also heard from senior industry leaders, who explored their own travel careers. ", "Jennifer Atkinson, chief executive of ITC Travel Group, urged up-and-coming travel professionals to “have the courage to do it [take on responsibility]” recalling how she took the helm at ITC aged 34.", "\n\n“When I first took the job I don’t believe I was any better than anybody else, I just happened to have an opportunity and seized it… now we are a £100 million business… I was brave enough to put my head above the parapet.”", "\n\nBrian Young, G Adventures’ managing director Europe, Middle East and Africa, told entrepreneurs to “take time to reflect on the down side”, recounting the failure of On Holiday Group in 2014, which he co-founded.", "\n\n“I didn’t realise just how much it would affect me personally,” he said, admitting his next immediate project subsequently “wasn’t fun for me at the time”.", "\n\nYoung advised delegates to choose carefully their next project after such a setback.", "\n\nMeanwhile, Jeannette Linfoot, managing director, Saga Tour Operating, urged delegates to be “tenacious”, explaining how she originally failed to secure a place on Thomson’s graduate scheme.", "\n\n“Travel’s not for the faint-hearted and it’s definitely not a 9-5 job,” she said. “", "People think it’s a party but it’s not – you have to work hard, but my god you will love it.”" ]

{ "pile_set_name": "Pile-CC" }

[ 0.005208333333333333, 0.00909090909090909, 0, 0.005154639175257732, 0, 0.012096774193548387, 0, 0, 0.015, 0, 0.014018691588785047, 0, 0, 0.015706806282722512, 0, 0 ]

[ "Q:\n\nGetting certain information from string\n\nI'm new to python as was wondering how I could get the estimatedWait and routeName from this string.", "\n{\n \"lastUpdated\": \"07:52\",\n \"filterOut\": [],\n \"arrivals\": [\n {\n \"routeId\": \"B16\",\n \"routeName\": \"B16\",\n \"destination\": \"Kidbrooke\",\n \"estimatedWait\": \"due\",\n \"scheduledTime\": \"06: 53\",\n \"isRealTime\": true,\n \"isCancelled\": false\n },\n {\n \"routeId\":\"B13\",\n \"routeName\":\"B13\",\n \"destination\":\"New Eltham\",\n \"estimatedWait\":\"29 min\",\n \"scheduledTime\":\"07:38\",\n \"isRealTime\":true,\n \"isCancelled\":false\n }\n ],\n \"serviceDisruptions\":{\n \"infoMessages\":[],\n \"importantMessages\":[],\n \"criticalMessages\":[]\n }\n}\n\nAnd then save this to another string which would be displayed on the lxterminal of the raspberry pi 2. ", " I would like only the 'routeName' of B16 to be saved to the string. ", " How do I do that?", "\n\nA:\n\nYou just have to deserialise the object and then use the index to access the data you want.", "\nTo find only the B16 entries you can filter the arrivals list.", "\nimport json\nobj = json.loads(json_string)\n\n# filter only the b16 objects\nb16_objs = filter(lambda a: a['routeName'] == 'B16', obj['arrivals'])\n\nif b16_objs:\n # get the first item\n b16 = b16_objs[0]\n my_estimatedWait = b16['estimatedWait']\n print(my_estimatedWait)\n\n" ]

{ "pile_set_name": "StackExchange" }

[ 0, 0.005698005698005698, 0.014492753623188406, 0, 0, 0.015873015873015872, 0.0035842293906810036 ]

[ "Tuberculosis in a prison population in Malawi.", "\nMuch concern has been expressed about the high prevalence of tuberculosis in prisons in industrialised countries. ", "Since there is almost no information from developing countries, we investigated the rate of pulmonary tuberculosis in a large prison in Malawi. ", "Between May and July, 1996; we carried out an active case-finding survey in Zomba Central Prison, Malawi, through the National Tuberculosis Control Programme. ", "We interviewed prisoners, and those with a cough of at least 1 week's duration were screened by sputum-smear microscopy. ", "If microscopy was negative, prisoners underwent chest radiography. ", "We offered HIV testing, with voluntary consent and counselling before and after tests, to all prisoners, whether positive or negative for pulmonary tuberculosis. ", "914 (70%) of 1315 prisoners were screened (905 men, nine women; mean age 30 years [SD 11]). ", "47 (5%) screened prisoners (all men) had pulmonary tuberculosis: 14 were taking antituberculosis treatment and 33 were undiagnosed at the start of the study; 18 were sputum-smear positive and 15 were sputum-smear negative. ", "16 (73%) of 22 prisoners with previously undiagnosed pulmonary tuberculosis and 30 (75%) of 40 prisoners with cough but no evidence of pulmonary tuberculosis were HIV seropositive. ", "In all prisoners, except one, symptoms of pulmonary tuberculosis had developed after they had entered prison. ", "We found a high rate of pulmonary tuberculosis in Zomba Central Prison, which suggests active transmission of tuberculosis. ", "As a result of this study, the National Tuberculosis Control Programme has implemented interventions in eight prisons in Malawi to improve tuberculosis control, including collection of health data, education of prisoners and clinical staff about tuberculosis, active screening of prisoners for pulmonary tuberculosis by sputum-smear microscopy, and active case-finding in the prisons." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0, 0, 0.012578616352201259, 0, 0, 0, 0, 0, 0, 0, 0, 0.0026041666666666665 ]

[ "No accelerator or brake pedals, either; self-driving is fully realized.", "\n\nA real-life prototype, left, and an artist's rendering, right, of Google's own self-driving car.", "\n\nGoogle has revealed it plans to build its own self-driving cars from the ground up, per an announcement from founder Sergey Brin at the Code Conference Tuesday. ", "The company revealed one such car to Recode, a highly compact two-seater without a steering wheel.", "\n\nThe prototype Google revealed differs from the Priuses and Lexuses in that they can't let humans take over the job of piloting; they are completely controlled by the onboard computer. ", "In addition to lacking a steering wheel, the Google-built car also has no accelerator, no brake, no mirrors, no glove compartment, and no soundsystem (your tiny smartphone speaker will have to do). ", "The cars are capped at a modest 25mph and are started and stopped by a button.", "\n\nIn a Q&A with Recode, head of the self-driving car project Chris Urmson stated that the car uses \"fault-tolerant architecture\" to minimize damage \"should something happen.\" ", "Urmson says that the the front end of the car is \"compressible foam\" and the windshield is flexible, which \"should do a much better job of protecting people if an accident should occur.\"", "\n\nGoogle gave no hints as to where the car was manufactured or a timeframe for official launch, so the project remains experimental. \"", "We’re going to learn a lot from this experience, and if the technology develops as we hope, we’ll work with partners to bring this technology into the world safely,\" states the official Google blog.", "\n\nWhile I can understand leaving out the steering wheel, brake, accelerator, etc. ", "since a typical self-driving scenario doesn't need it, what about when you just want to drive it a short distance, like moving it out of the way if it's blocking a car in a driveway? ", "How would you tell it to park in an empty spot once you get to your destination? ", "I'm wondering if they have ways of dealing with that with this vehicle or if they simply aren't even concerning themselves with these questions at this point.", "\n\nAnother scenario: If you're driving along and suddenly want to stop for something (you spot a friend, an accident nearby, etc.), ", "how would you do that?" ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0.01020408163265306, 0.006134969325153374, 0.01020408163265306, 0.005376344086021506, 0, 0, 0.011428571428571429, 0.005376344086021506, 0, 0.005050505050505051, 0, 0, 0, 0, 0, 0 ]

[ "All good points for a larger project,\nyet less interesting for a personal cluster when the budget is small\nand the personal desktop and personal work environment is Ubuntu. ", "I do\nsuspect that in 18 months or so the original poster will be looking to\nupdate his environment and your good summary will then be apropos.", "\nEspecially the bit \"a source based \"ports-like\" tree for building scientific\npackages, and then we integrate with Environment Modules to make them\navailable for the users.....\". ", "See module-assistant in Ubuntu and friends.", "\nI just deleted a longish comparison of GenToo and British sports cars. ", "Since\nthe OP is not a mechanic and does not have a staff, mechanic or chauffeur to\nmaintain his roadster. ", "I suspect he needs a Toyota, Ford or Chevy with an\nautomatic transmission that he can maintain himself (for now).", "\nOh wait I almost retyped it all.... ;-)\nOn Mon, Oct 12, 2009 at 11:50:10AM -0700, Greg Kurtzer wrote:\n> On Sun, Oct 11, 2009 at 1:29 PM, Nifty Tom Mitchell\n> <niftyompi at niftyegg.com> wrote:\n> > On Mon, Oct 05, 2009 at 12:02:11PM +0200, Tomislav Maric wrote:\n> >> Nifty Tom Mitchell wrote:\n> >> > On Sun, Oct 04, 2009 at 01:08:27PM +0200, Tomislav Maric wrote:\n> >> >> Mark Hahn wrote:\n> >> >>>> I've seen Centos mentioned a lot in connection to HPC, am I making a\n> >> >>>> mistake with Ubuntu??", "\n> >> >>> distros differ mainly in their desktop decoration. ", "for actually\n> >> >>> getting cluster-type work done, the distro is as close to irrelevant\n> >> >>> as imaginable. ", "a matter of taste, really. ", "it's not as if the distros\n> >> >>> provide the critical components - they merely repackage the kernel,\n> >> >>> libraries, middleware, utilities. ", "wiring yourself to a distro does\n> >> >>> affect when you can or have to upgrade your system, though.", "\n> >> >>>\n>> I seem to not have the original sources to this thread, but this is\n> something that I thought I should chime in on.", "\n>> The underlying components that make up a distribution are in-fact an\n> important component to an HPC system in its entirety. ", "There are many\n> reasons for this, but I will focus on just a few that I hope don't\n> strike too much of a religious chord with people while at the same\n> time letting me rant a bit. ;-)", "\n>>>> 1) HPC people are quite familiar with building their scientific apps\n> with optimized compilers and libraries. ", "If an application is linking\n> against any OS libraries (yes, including the C library) it would\n> probably make sense to make sure those have been compiled with an\n> optimal build environment. ", "Most distributions do not do this, as for a\n> single standalone system the results may or not even be noticeable. ", "I\n> have been part of large benchmark projects to evaluate the differences\n> of the distributions. ", "In a nutshell, differences become more obvious\n> at scale.", "\n>> 2) Distributions focused on non-HPC targets may not include tools,\n> libraries or even functions that would be beneficial for HPC. ", "And in\n> addition to that, they may not be included in a way that makes it very\n> usable. ", "For example, Just because a distribution contains a package\n> does not mean that is what people should use. ", "For example, using a\n> distribution supplied version of Open MPI would be an injustice to the\n> majority of cluster users, but many distributions consider themselves\n> HPC ready because they have some HPC capable libs.", "\n>> It is more important to have a solution for creating a suitable HPC\n> environment. ", "For example, in Caos NSA the core OS is RPM based but we\n> also utilize a source based \"ports-like\" tree for building scientific\n> packages, and then we integrate with Environment Modules to make them\n> available for the users. ", "So one can do:\n>> # cd /usr/src/cports/packages/openmpi/1.3.3\n> # make install COMPILERS=intel\n> # make clean\n> # make install COMPILERS=gcc\n> # su - user\n> $ module load openmpi/1.3.3-intel\n> $ mpicc -show\n> icc ........\n> $ module unload openmpi\n> $ module load openmpi/1.3.3-gcc\n> $ mpicc -show\n> gcc ........\n>>> 3) HPC distributions should be focused on being lightweight and\n> efficient. ", "Bloat free stateless environments are important for keeping\n> node operating systems quiet and supportive of HPC code. ", "Lightweight\n> and bloat free does not mean an ancient and featureless core\n> environment either.", "\n>> 4) Even the kernel for most distributions is tuned for desktop use\n> which tries to give the fairest share of CPU time to all processes\n> (obviously not HPC supportive).", "\n>> 5) Clusters are not worth BETA quality code. ", "Unstable environments\n> with no long term plans for upstream support makes it a ridiculous\n> solution for anybody trying to build a production environment. ", "The\n> number of unsuitable solutions that we had to \"rescue\" because they\n> were running Fedora and totally unmaintainable by the people that\n> integrated them is just silly.", "\n>>> It really is a shame that religion plays such a big part of what OS\n> someone would use because each OS (even non-Linux) are good for\n> certain things. ", "I can understand wanting to leverage an economy of\n> scale with a homogenous environment, but there is a particular point\n> where the economy of scale is no longer justified when shoe-horning a\n> non-suitable solution onto a cluster. ", "Where that line is really\n> depends on the admins, users, the size of the system, and what they\n> baseline their benchmark for success at.", "\n>> Just my $0.02 for what it's worth.", "\n>> Greg\n>>> --\n> Greg M. Kurtzer\n> Chief Technology Officer\n> HPC Systems Architect\n> Infiscale, Inc. - http://www.infiscale.com\n--\nT o m M i t c h e l l\nFound me a new hat, now what?", "\n_______________________________________________\nBeowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing\nTo change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf" ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0, 0.00558659217877095, 0, 0, 0.009433962264150943, 0.02654867256637168, 0.022044088176352707, 0, 0, 0, 0, 0, 0, 0, 0.005376344086021506, 0, 0, 0, 0, 0, 0.007407407407407408, 0, 0, 0.0045871559633027525, 0.011494252873563218, 0.008771929824561403, 0, 0, 0, 0, 0, 0, 0.005747126436781609, 0, 0, 0, 0, 0.010869565217391304, 0.01702127659574468 ]

[ "Whitaker test: differentiation of obstructive from nonobstructive uropathy.", "\nThe Whitaker test, a urodynamic study, combined with antegrade pyelography has been used recently to evaluate persistent upper urinary tract dilatation after operative correction of obstruction. ", "This test will differentiate patients with residual or recurrent obstruction from those with dilatation secondary to permanent changes in the musculature. ", "It is useful in evaluating patients with questionable ureteropelvic or ureterovesical junction obstruction, or primary defects in the ureteral musculature, such as prune-belly syndrome. ", "It also will establish when urinary diversion safely can be discontinued in postoperative patients. ", "Percutaneous puncture of the renal pelvis is performed or an indwelling nephrostomy tube is used. ", "The upper urinary tract is then perfused at a constant rate of 5--10 ml/min with saline or diluted contrast media, and a serial pressure recording is made in the renal pelvis and bladder. ", "The high flow rate used will be tolerated easily in a nonobstructed system without a progressive rise in renal pelvic pressure. ", "In obstructed systems abnormally high pressure above 12 cm water or a constant rise in pressure will be recorded. ", "Videotaping of ureteral peristalsis and spot films of the upper urinary tract complete the evaluation." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0.00510204081632653, 0, 0, 0, 0, 0, 0, 0, 0 ]

[ "This is listing of dumped prototypes that are documented on SNES Central and differ from the final version. ", "To see undumped prototypes, or prototypes that are identical to final, please go to the specific game page. ", "I'm including information here so that it may eventually be used for emulators such as Higan. ", "This page also includes dumped unreleased games, but I recommend going to the unreleased games section for more information.", "\n\nHere is a guide for the table below. ", "For the \"dump status\", games that have been dumped by someone known are set as \"verified\", known hacked ROM images are set as \"hacked\", incomplete or damaged (bad) dumps are listed as such, and \"unverified\" means that I don't know the original dumper and cannot verify that the game has not been tampered with. ", "Unfortunately, many prototypes were leaked many years ago on BBSes or on the early Internet and the original cart may be long gone. ", "The board type is the PCB that should give the correct memory mapping for emulation. ", "In general, I set it to be the same as the commercial release. ", "Note that if you use incorrect memory mapping, the game may not work correctly if the internal header information is incorrect. ", "For instance, if you try playing the prototype of The Flintstones: Treasure of Sierra Madrock with an emulator, it will automatically map it to have a battery, which will trigger copy protection that makes the game nearly unplayable.", "\n\nThe SHA-256 hashes below were generated using NSRT 3.4. ", "The hashes may differ from what you would get from other SHA256 generators, since NSRT heuristically determines the structure of the ROM image to account for copier headers, for instance. ", "Sometimes there are false positives, which leads to a SHA256 that is different from a file agnostic program. ", "In the future, I may decide to change to a more traditional SHA256 generator." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0, 0.010638297872340425, 0, 0, 0.003215434083601286, 0.007575757575757576, 0, 0, 0, 0, 0, 0.005319148936170213, 0, 0 ]

[ "Diffusion Monte Carlo simulations of gas phase and adsorbed D2-(H2)n clusters.", "\nWe have computed ground state energies and analyzed radial distributions for several gas phase and adsorbed D2(H2)n and HD(H2)n clusters. ", "An external model potential designed to mimic ionic adsorption sites inside porous materials is used [M. Mella and E. Curotto, J. Phys. ", "Chem. ", "A 121, 5005 (2017)]. ", "The isotopic substitution lowers the ground state energies by the expected amount based on the mass differences when these are compared with the energies of the pure clusters in the gas phase. ", "A similar impact is found for adsorbed aggregates. ", "The dissociation energy of D2 from the adsorbed clusters is always much higher than that of H2 from both pure and doped aggregates. ", "Radial distributions of D2 and H2 are compared for both the gas phase and adsorbed species. ", "For the gas phase clusters, two types of hydrogen-hydrogen interactions are considered: one based on the assumption that rotations and translations are adiabatically decoupled and the other based on nonisotropic four-dimensional potential. ", "In the gas phase clusters of sufficiently large size, we find the heavier isotopomer more likely to be near the center of mass. ", "However, there is a considerable overlap among the radial distributions of the two species. ", "For the adsorbed clusters, we invariably find the heavy isotope located closer to the attractive interaction source than H2, and at the periphery of the aggregate, H2 molecules being substantially excluded from the interaction with the source. ", "This finding rationalizes the dissociation energy results. ", "For D2-(H2)n clusters with n≥12, such preference leads to the desorption of D2 from the aggregate, a phenomenon driven by the minimization of the total energy that can be obtained by reducing the confinement of (H2)12. ", "The same happens for (H2)13, indicating that such an effect may be quite general and impact on the absorption of quantum species inside porous materials." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0.01282051282051282, 0.007194244604316547, 0.007352941176470588, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.006535947712418301 ]

[ "FISH analysis of fetal nucleated red blood cells from CVS washings in cases of aneuploidy.", "\nIn chorionic villus sampling (CVS) the chromosome analysis is inconclusive in 1-2% of the samples. ", "In many cases follow-up amniocentesis is performed. ", "Fetal nucleated red blood cells (FNRBCs) are present in washings of chorionic villus samples. ", "We wanted to establish whether analysis of these true fetal cells, using fluorescence in situ hybridization (FISH), could support the CVS karyotype. ", "We analysed washings of first trimester chorionic villi from non-mosaic 45,X (n=6) and full trisomy 18 cases (n=7). ", "FNRBCs were identified by immunostaining and FISH was performed with chromosome-specific probes for X, Y and 18. ", "In all 13 samples FNRBCs were present (between 4 and 30 cells per sample). ", "Five cases of monosomy X showed one X signal in 89-100% of the nuclei; in the other case 50% of the nuclei displayed one signal. ", "In the trisomy 18 cases three spots were seen in 60-100% of the cells. ", "The CVS aneuploidy was confirmed in FNRBCs in all samples, so FISH on FNRBCs can be used in cases of non-mosaic numerical chromosomal abnormalities. ", "This test can confirm a CVS diagnosis of monosomy X or trisomy 18 and thus minimize the risk for false-positive diagnoses. ", "An additional invasive test may be prevented." ]

{ "pile_set_name": "PubMed Abstracts" }

[ 0, 0.01, 0, 0, 0.006711409395973154, 0, 0, 0, 0, 0, 0.006711409395973154, 0, 0 ]

[ "MLB News\n\nFreese won't rub it in with ex-Rangers on Angels\n\nBy\nAlden GonzalezMLB.com\n\nTEMPE, Ariz. -- Six players from that epic seven-game 2011 World Series are now in camp with the Angels, with three from the Rangers (Josh Hamilton, C.J. Wilson, Yorvit Torrealba) and three from the Cardinals (Albert Pujols, David Freese, Fernando Salas).", "\n\nThe hero from that Fall Classic promises that when it comes to the former Rangers who are now his teammates, all that is in the rearview mirror.", "\n\nTEMPE, Ariz. -- Six players from that epic seven-game 2011 World Series are now in camp with the Angels, with three from the Rangers (Josh Hamilton, C.J. Wilson, Yorvit Torrealba) and three from the Cardinals (Albert Pujols, David Freese, Fernando Salas).", "\n\nThe hero from that Fall Classic promises that when it comes to the former Rangers who are now his teammates, all that is in the rearview mirror.", "\n\n\"They're Angels,\" Freese, who batted .348/.464/.696 as the 2011 World Series MVP, said upon arriving to Angels camp on Sunday morning. \"", "I don't even really think about the World Series too much anymore.\"", "\n\nFreese carried a historic postseason performance into a highly productive 2012 season that saw him bat .293/.372/.467. ", "Last year, though, he bruised his tailbone flipping over a railing in Spring Training, started the season on the disabled list, lost a lot of his range at third base, saw his slash line drop to .262/.340/.381 in the regular season, then posted a .526 OPS in the Cardinals' run to the World Series.", "\n\nHe's fully healthy now, with a fresh state of mind.", "\n\n\"There's an excitement, an edge, I have coming here,\" Freese said. \"", "It's going to help mentally.\"" ]

{ "pile_set_name": "Pile-CC" }

[ 0.03519061583577713, 0.00684931506849315, 0.0311284046692607, 0.00684931506849315, 0.014492753623188406, 0, 0, 0.006734006734006734, 0, 0, 0 ]

[ "On Air Force One, President Trump said again that there were bad people on the other side of the white supremacy protests in Charlottesville, Virginia." ]

{ "pile_set_name": "OpenWebText2" }

[ 0.006622516556291391 ]

[ "psychiatrist near me kaiser - An Overview\n\npsychiatrist near me kaiser - An Overview\n\nKaiser is really fairly handy with regard to encouraging people today to implement their teams, they may have dozens of these. ", "You can get a person therapist, as well as an individual psychiatrist. ", "The psychiatrists I've achieved with often were being basically pretty generous with their time, and have been incredibly perceptive. ", "If you want long lasting talk therapy, I'd recommend maxing out your lined Kaiser selections, then on the lookout elsewhere.", "\n\nShould your browser won't request you, test these measures: At the top within your Chrome window, near the world wide web address, click the green lock labeled Safe.", "\n\nI am so sorry regarding your step mom. ", "I am so happy you have been in a position to get help from Kaiser after you necessary it.", "\n\nBigger managed meds like cymbalta, adderal and so on are minor tougher. ", "Exclusive get goods we don't have really are a nightmare however. ", "Excellent luck For anyone who is coming in just after five:30pm after the standard Physicians leave. ", "You'll be able to appear into my pharmacy right up until nine:00pm and yell at me, Everybody does.", "\n\nAt the beginning I blew it off and did not Visit the group, but my counselor insisted, so I went. ", "I thought these folks are depressed which is not me? ", "But you determine what I had been grieving and it designed me depressed and also the 12 week system genuinely helped me. ", "It was the moment per week for an hour or so in addition to a fifty percent. ", "And I had been seriously in opposition to The full \"team therapy\" issue in the get go, but it really genuinely did aid me Get well.", "\n\nIf you need weekly visits, the Pacific Medical Ctr psychological college in SF is website here very inexpensive, I think It is like $thirty for an hour along with a 50 %, my mates who have absent there have had seriously very good luck with fantastic therapists. ", "They may be pupils, nonetheless. ", "There are a selection of other free and sliding scale therapy choices in SF also.", "\n\nEnter the shape, color, or imprint of your prescription or OTC drug. ", "Our pill identification Instrument will Show shots which you can Examine to your tablet.", "\n\nSo a few of you could say \"Kaiser Sucks\" but its improved that possessing \"no\" protection, which loads of individuals are experiencing these days.", "\n\n0 assessments I recently attempted to get assistance at Kaiser because of melancholy about a Serious suffering challenge. ", "I had to attend two weeks for an appointment. ", "It is a complete crapshoot simply because you are unable to investigation your therapist. ", "I had been Blessed, the MFT I noticed was fantastic, but their caseload is so significant that she will be able to only see me each individual 6 months. ", "They want you to definitely visit \"group\" but even the find more info MFT claimed These groups are so overloaded that at many of them you merely go and listen to the therapist read through the material to you. ", "I questioned for an appointment using a psychiatrist and additional reading had to attend four months.", "\n\nNot to reduce your knowledge, but have you tried out acupuncture or hypnotherapy? ", "My Mate's Trainer was bedridden For a long time due to extreme again agony, and she managed to cope with her discomfort via a incredibly highly regarded acupuncturist in China town.", "\n\nalso, um, many thanks for permitting us know that you'll be not crazy enough for psychological wellness treatment. ", "Contrary to popular belief, \"crazy\" folks are not the one individuals who get Positive aspects from psychotherapy.", "\n\n221 opinions Kaiser sucks for mentalhealth protection. ", "You basically must be suicidal, then they can put you in an Intensive Outpatient System.", "\n\nA clinical psychologist uses psychotherapy as well as other counseling expertise to improve psychological and mental wellness. ", "A Psychiatrist is really a health care physician who performs psychotherapy but also can prescribe drugs." ]

{ "pile_set_name": "Pile-CC" }

[ 0.004694835680751174, 0, 0, 0.008064516129032258, 0.005988023952095809, 0, 0.011235955056179775, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.0037735849056603774, 0, 0, 0, 0, 0, 0.008064516129032258, 0, 0, 0.013071895424836602, 0.004761904761904762, 0, 0, 0.0055248618784530384, 0, 0, 0.017543859649122806, 0.011363636363636364, 0, 0 ]

[ "Incentives matter; everything else is conjecture. ", "That lesson was taught to me by Donald Marron, a brilliant economist who I worked for at the Joint Economic Committee. ", "His point was profound, yet simple: if you want to figure out the end effect of any policy proposal, examine how it will change people’s incentives.", "\n\nMarron’s theory is especially relevant when it comes to Obamacare, which drastically changed incentives for workers and employers throughout the country. ", "Because of the requirements that companies of a certain size offer health care coverage to full-time employees or face stiff government fines, many employers have started to either drop coverage altogether (the fines are cheaper) or reduce the amount of full-time employees to avoid the mandates.", "\n\nThe following story from SFGate.com shows how Obamacare also warps the incentives for workers:\n\nPeople whose 2014 income will be a little too high to get subsidized health insurance from Covered California next year should start thinking now about ways to lower it to increase their odds of getting the valuable tax subsidy.", "“If they can adjust (their income), they should,” says Karen Pollitz, a senior fellow with the Kaiser Family Foundation. “", "It’s not cheating, it’s allowed.", "”Under the Affordable Care Act, if your 2014 income is between 138 and 400 percent of poverty level for your household size, you can purchase health insurance on a state-run exchange (such as Covered California) and receive a federal tax subsidy to offset all or part of your premium.", "For people in their early 60s, “it’s a huge cliff,” going from 401 to 400 percent of poverty, Pollitz says. ", "That’s because insurers can still charge older people more than younger ones.", "\n\nBecause of the clumsy manner in which Obamacare defines who gets subsidies and who doesn’t, there is a huge “cliff effect” at 400% of the federal poverty line (FPL). ", "If you make a penny more, you won’t receive health care subsidies. ", "If you make a penny less, you could receive thousands in subsidies. ", "The result? ", "Rational people who are right around the 400% FPL threshold will choose to work less in order to increase their income after paying for health insurance.", "\n\nIn California, a couple earning $64,000 a year would not qualify for health care subsidies. ", "A bronze plan for them through Kaiser would cost them about $1,300 each month, or $15,600 a year. ", "But if that same family earned just $2,000 less, it would qualify for over $14,000 in annual health care subsidies, dropping their premiums for that same Kaiser plan to less than $100 per month.", "\n\nThe after-insurance income for the couple earning $64,000 is $48,400. ", "The after-insurance income for the couple making $62,000, which is just under the 400% FPL threshold, is $60,800.", "\n\nIt therefore makes no sense for a rational couple on the FPL bubble to work a little harder to earn a little extra money, because Obamacare severely punishes them for their work ethic. ", "That couple ends up significantly worse off for having tried to earn more money. ", "And we’re not talking about an effective marginal tax rate of 40 or 50 or even 90 percent here. ", "If you’re on the FPL bubble, the effective marginal tax rate is in the neighborhood of 10,000 percent. ", "That next dollar you earn could result in you losing over $10,000 in health care benefits.", "\n\nJust as Obamacare created huge incentives for employers to either dump health plans, dump hours, or dump workers, it also created horribly perverse incentives for otherwise hard-working people to not work so hard. ", "The result? ", "Fewer jobs, lower incomes, and less economic growth." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0.01680672268907563, 0, 0.00641025641025641, 0, 0, 0.01639344262295082, 0, 0, 0.009259259259259259, 0, 0.005952380952380952, 0, 0, 0, 0.006535947712418301, 0, 0.01020408163265306, 0.005154639175257732, 0, 0.008849557522123894, 0.0053475935828877, 0, 0, 0.009708737864077669, 0, 0, 0, 0 ]

[ "Dateline Disneyland\n\nWelcome to the final edition of Dateline Disneyland for 2010! ", "We hope you had a fantastic Christmas holiday and are ready to ring in the new year! ", "But before we can welcome 2011, it's time to take a look back at a selection of the major happenings around the Disneyland Resort over the past year. ", "From a major Rivers of America overhaul to Disneyland's 55th anniversary, Disneyland has seen a lot of activity this year. ", "Yet, with loads of construction, the new World of Color water-and-light show, and after-dark dance parties; nothing compares to the major goings-on at Disney California Adventure. ", "It's been quite the year, so join me and let's take a look back at 2010 at the Disneyland Resort.", "\n\nWELCOME 2010!", "\nDisneyland celebrated the arrival of 2010 with its usual array of dance parties, entertainment and midnight fireworks in both parks.", "\n\nSTROLLER SHOPPED\nA new central stroller and wheelchair rental location opened in early January in the Resort's Main Entrance Plaza. ", "The new stroller shop eliminated the rental locations in both parks, basing rentals out of one central location for the Resort theme parks.", "\n\nAfter the walls came down:\n\nBYE-BYE HONEY, HELLO CAPTAIN\nTomorrowland's Magic Eye Theater closed at the start of the new year evicting the tired \"Honey, I Shrunk the Audience\" show to make way for the triumphant return of the 1986 \"Captain EO\" 3D film starring Michael Jackson.", "\n\nCaptain EO didn't open its theater doors until late February, but the return of the retro 3D show once again made the Magic Eye Theater a popular destination in Tomorrowland.", "\n\nBRIDGING THE FRONTIER\nIn February, the gateway to Frontierland got a new bridge with higher safety rails. ", "The new bridge is themed well enough, but the old bridge was less obtrusive, visually. ", "Some will disagree that a simple bridge isn't all that important, but the new lawyer-friendly bridge seemed to diminish the simple charm of the park's central plaza.", "\n\nMeanwhile, the waterways surrounding the entrance to Frontierland were refurbished, resulting in a waterway that looked more like a pond from a miniature golf course rather than a stream you'd encounter in the American frontier.", "\n\nThe theming blunder was quickly fixed, the cement lip was removed and a excellently themed alternative was installed.", "\n\nTWO BROTHERS TUNEMAKERS\nIn March, Disney Legends Richard and Robert Sherman were honored with a window on Main Street.", "\n\nOver in Fantasyland, an attraction famous for a Sherman Brothers song, \"it's a small world,\" got a new marquee.", "\n\nSHOPPING SHUFFLE\nIn April, New Orleans Square's Jewel of Orleans closed after third-party retailer Dianne's Estate Jewelry vacated the location after 12 years of business in the park. ", "The closure was due to the economic downturn, but the loss of the antique jeweler was certainly a loss to the character and makeup of the New Orleans Square area.", "\n\nThe shop reopened in May as an upscale women's accessory and handbag shop run by Disney.", "\n\nIn Fantasyland, the Gepetto's Candy Shoppe opened after the retail location sat empty for several months.", "\n\nThe candy shop was short lived, however, as the space became home to the character meet-and-greet for Rapunzel and Flynn Rider from Disney's animated film Tangled in October.", "\n\nRIVER RESIDENTS\nIn early May, Frontierland's Rivers of America emerged from a five-month refurbishment with new scenery and wildlife vignettes.", "\n\nDisneyland visitors will soon be able to see Native Americans scouting the territory and a keel boat in front of a cabin on Tom Sawyer Island for the first time.", "\n\nThe theme park’s Rivers of America, the main waterway surrounding the island, reopened Saturday after a four-month rehabilitation project. ", "Disneyland crews drained the concrete-bottomed river starting Jan. 4 so it could complete upgrades, including the first new scenes on the shoreline since 1991.", "\n\nThe river began refilling earlier this week with recycled water just like was done with the water for the lagoon at Disney’s California Adventure.", "\n\nSome of the new scenes included a new resident in the settler’s cabin on the island, Mike Fink from Walt Disney’s television mini-series, “Davy Crockett.”", "\n\nA small herd of deer were spotted near a new stream tumbling down the rocks from the tracks for the Disneyland Railroad.", "\n\nThe Columbia Sailing Ship was parked at the dock as crews put the finishing touches on several of the scenes and attraction operators made sure the ship and the rafts that take guests to Tom Sawyer Island were ready to go.", "\n\nThe waterway now is divided into four U.S. river regions: the Mississippi, the Columbia, the Potomac and the Rio Grande. [ ", "FULL STORY ]\n\nCELEBRATING 55 YEARS\nIn late May, the Disney Gallery opened its \"Day One Disneyland\" exhibit celebrating the park's 55th anniversary.", "\n\nDespite the milestone year, the park's July 17th anniversary date was rather subdued. ", "Still, a pleasant and nostalgic rededication ceremony in Town Square was packed with park fans, and a special merchandise release sold out of nearly everything.", "\n\nA \"flash mob\" of park Cast Members entertained guests after the anniversary rededication.", "\n\nThe Main Street Cinema showed newly-found footage of the park's early days for a limited time in celebration of the 55th anniversary.", "\n\nIn the fall, D23 - the official Disney fan club, hosted Destination D: Disneyland '55, a celebration of the park's history featuring numerous Disney celebrities, artists and Imagineers.", "\n\nImagineers Marty Sklar and X Atencio\n\nImagineer Tom Fitzgerald discusses Star Tours\n\nSTAR TOURED\nJuly 26th marked the closing of Star Tours as the attraction undergoes major changes and upgrades in preparation of its reopening this spring.", "\n\nThe attraction is set to feature new characters and experiences when it reopens in 2011.", "\n\nALICE IN BLUNDERLAND\nAt the end of July, Fantasyland's Alice in Wonderland abruptly closed and safety measures were taken to prevent maintenance workers from falling from the attraction's elevated outdoor portion of track after a Cal/OSHA employee noticed a potential safety hazard.", "\n\nThe attraction reopened with ugly scaffolding covered in green tarps holding up safety walkways. ", "The views on and off the attraction were ruined and much of the charm was lost. ", "A permanent fix is expected to come in 2011.", "\n\nCALIFORNIA CHANGES\n2010 brought big changes to Disney California Adventure, including a new logo, new attractions and lots of construction!", "\n\nTROLLEY TRACKING\nConstruction got underway on the new Red Car Trolley tracks at the start of 2010 in the Hollywood Pictures Backlot. ", "Construction walls announced the new attraction and an open date set for May 2012. ", "Construction was spread out through several phases down Hollywood Boulevard down to the Twilight Zone Tower of Terror throughout the year, including the opening of a temporary pathway between the Tower of Terror and A Bug's Land, which was later closed and reopened as a permanent connector in August.", "\n\nWELCOME TO PARADISE\nIn April, Paradise Park opened to guests after over a year of construction. ", "The new area, completely rethemed and rebuilt to accommodate crowds gathering to view the World of Color show was a major step in the overall re-Imagining of the park.", "\n\nTOY STORY 3 MANIA\nIn May, Toy Story Midway Mania got its first change, with the removal of Bo Peep's Ba-a-loon Pop and the introduction of Rex and Trixie's Dino Darts. ", "The change was the first time the attraction saw a new game added and the first time characters from Toy Story 3 were seen in the attraction.", "\n\nA SWINGIN' SYMPHONY\nBy late May the former Orange Stinger reopened as the Silly Symphony Swings, bringing some charm to Paradise Pier after ditching the tacky theme the ride opened with in 2001.", "\n\nCALIFORNIA, REBRANDED\nAs the Silly Symphony Swings opened in May, Disney unveiled a new logo for the California park, ditching the previous two park logos:\n\n...and introducing a new, more-cartoony logo that plays on style of opening titles from vintage Mickey Mouse cartoons. ", "The new logo also dropped the possessive form of \"Disney\" from the park name, changing it to \"Disney California Adventure\"\n\nA WONDERFUL WORLD OF COLOR\nWith a new logo and the opening of the Silly Symphony Swings ride, California Adventure was approaching the debut of its new \"World of Color\" show. ", "A commercial promoted the new show to TV audiences as park guests saw the work crews testing the show's fountains during park hours, and Resort hotel guests saw sneak peeks of dress rehearsals after dark from their hotel room windows.", "\n\nFinally, after over a year of construction and anticipation, the show finally debuted with a cavalcade of Disney celebrities in attendance and instantly became a crowd favorite.", "\n\nGLOW FEST & ELECTRONICA\nIn an effort to give guests waiting for World of Color something to do in California Adventure until it was time to see the show, Disney had a third-party event team draw up Glow Fest, a neon-colored dance party that took over the Hollywood Pictures Backlot with a giant dance party and free-standing bars. ", "The event proved to be wildly successful and became a popular attraction on its own for park guests throughout the summer.", "\n\nThe success of Glow Fest led Imagineers to create their own after-dark dance party following the summer run of Glow Fest. ", "The Disney-created event utilized the new Tron: Legacy film as a basis and the end result was ElecTRONica, which included extra attractions including an extended 3D preview of the new film and a recreation of Flynn's Arcade along with dancing and bars.", "\n\nWith ElecTRONica came World of Color's first new scene. ", "The \"Troncore,\" a new show scene featuring music and visuals from \"Tron: Legacy\" debuted in late October.", "\n\nCONSTRUCTING CARS LAND\nConstruction sped up on Cars Land in 2010 and in May the first test car arrived at the upcoming Radiator Springs Racers attraction as crews worked on testing the ride system and track.", "\n\nSoon, the massive Radiator Springs mountain range began to take shape and buildings along the land's main street started going up.", "\n\nIn September, the old Bountiful Valley Farm area of the park closed and most of the area's buildings were bulldozed to make way for more of Cars Land.", "\n\nIn late October, Imagineers celebrated the \"topping out\" of Cars Land construction, marking the highest point of vertical construction on the project.", "\n\nThe Disney Parks Blog announced the construction milestone:\n\n“Topping out” is a traditional construction event among construction workers and engineers that takes place when the highest piece of steel is installed on the frame of a building. [ . . . ]", "\n\nAs part of the tradition, the final beam is painted white and signed by the ironworkers, steel design engineer and construction management. ", "This piece of steel will be part of the Pinnacle of the Cadillac Range, representing a 1959 Cadillac, whose tail fin was also known as the Pinnacle.", "\n\nSeveral key people were on hand to explain the significance of this milestone, including Vice President of Disney California Adventure park, Mary Niven, Walt Disney Imagineer Kathy Mangum, and Walt Disney Imagineer Jim Kearns.", "\n\nMary Niven described the event as a huge milestone for Disney California Adventure park. “", "Literally and metaphorically, we’ve reached the other side of the mountain,” she said.", "\n\nLITTLE MERMAID CONSTRUCTION\nRapid progress marked the construction of the upcoming Little Mermaid attraction, with the attraction's show building being constructed practically overnight in the empty space left from the former Golden Dreams attraction.", "\n\nThe new Little Mermaid attraction is set to open this spring.", "\n\nDECONSTRUCTING THE MALIBOOMER\nParadise Pier's looming Maliboomer thrill ride closed permanently in early September as crews began dismantling the attraction on September 7. ", "By the end of October there was nothing left of the ride. ", "Current plans for the space left by the ride call for a shady park-like area with character meet-and-greets.", "\n\nCONSTRUCTION MADNESS\nAlso in September, Paradise Pier's Pizza Oom Mow Mow and the surrounding area closed to begin retheming.", "\n\nThe rest of the tacky S.S. rustworthy play area was finally removed.", "\n\nThe long-closed Burger Invasion finally went under the knife\n\nAnd in October, the nearby Mullholland Madness rollercoaster closed to be rethemed as Goofy's Sky School, which is expected to open in the spring.", "\n\nSUN-SHINED\nMid-September marked the start of demolition in Sunshine Plaza, DCA's main entrance area. ", "Construction crews began removing the park's giant tile murals that were intended to make park guests feel as though they were stepping into a giant postcard.", "\n\nAlso removed was the sun sculpture and fountain and the Sunshine Plaza planters.", "\n\nLOOKING AHEAD\nWhile 2010 was a busy year for DCA, 2011 will be even busier as work gets underway on Buena Vista Street and construction ramps up on Cars Land. ", "Brady MacDonald at the LA Times has a timeline of upcoming projects in California Adventure.", "\n\nAs part of an ongoing, $1.1 billion makeover, the sweeping and extensive reimagineering will wall off 4 acres of the main entrance between the C-A-L-I-F-O-R-N-I-A letters, Sunshine Plaza, Grizzly Peak and Hollywood Studios Backlot. ", "The remade entryway, dubbed Buena Vista Street, will be designed to evoke the era when a young Walt Disney first arrived in Los Angeles.", "\n\nAfter the winter holiday season ends, construction walls will go up around the iconic C-A-L-I-F-O-R-N-I-A letters Jan. 4 so that crews can turn the plaza area into Streamline Moderne turnstiles modeled after the former Pan-Pacific Auditorium in Los Angeles.", "\n\nThe most disruptive phase of the project will start in August, when retheming work will begin on the shops just inside the entrance. ", "On Aug. 1, Engine-Ear Toys, Bu-r-r Bank Ice Cream and Baker's Field Bakery will close on the right side of the entrance. ", "Then on Aug. 29, the Greetings from California gift shop on the left side of the entrance will close. ", "If all goes according to plan, the new Pan-Pacific-style turnstiles will also open Aug. 29.", "\n\nBut construction will force Disney to route visitors through a 500-foot-long backstage pathway running behind Soarin' Over California. ", "The temporary entrance/exit route is expected to remain in place through spring 2012. ", "Another pathway carved through the Sunshine Plaza construction zone will allow visitors to travel between Condor Flats and Hollywood Studios Backlot.", "\n\nThe undertaking is so massive and unprecedented that DCA Vice President Mary Niven has taken to a Web video in an effort to convince visitors the Anaheim theme park will remain open for business once the front gate turns into a construction zone.", "\n\n\"While we're adding all these new attractions, our guests' favorite attractions are still open,\" said Niven, who faces the daunting challenge of building and operating a theme park at the same time.", "\n\nIndeed, all major attractions and shows will remain open during construction.", "\n\nBy spring 2011, virtually all the areas of Paradise Pier currently behind construction walls will reopen to the public – including the Voyage of the Little Mermaid dark ride, the Goofy's Sky School wild mouse coaster, the Paradise Garden Grill and Boardwalk Pizza and Pasta restaurants, the Corndog Castle take-out stand and the SeaSide Souvenir shop.", "\n\nThe 12-acre Cars Land addition is expected to make its debut with three new rides in spring 2012.", "\n\nThe ongoing expansion, which some have derisively dubbed \"Disney's Construction Adventure,\" has proven to be a continuing nightmare for DCA planners. ", "They were forced to ship the Main Street Electrical Parade to the Magic Kingdom when earlier construction shrank the parade route and cancel the California Food & Wine Festival for 2011 and 2012 when logistical hurdles became evident.", "\n\nIn addition, the Pixar Play Parade will go on hiatus after Jan. 3, with the parade characters continuing to appear in pop-up \"dance parties\" throughout the park. \"", "The Incredibles\" mini float will begin appearing at various locations throughout the park on Jan. 7, with \"Toy Story\" and \"Monsters, Inc.\" pop-up parties making their debuts over the ensuing weeks. ", "The full Pixar parade is expected to return in November.", "\n\nThe Playhouse Disney theater will also close after Jan. 23 for conversion into the Disney Junior theater in spring. ", "The rethemed theater will feature a new show segment based on the upcoming Disney Channel cartoon \"Jake and the Never Land Pirates.\"", "\n\nAdditionally, the Blue Sky Cellar preview center will add a more elaborate display on the Little Mermaid ride beginning on Jan. 27. [ ", "FULL STORY ]\n\nBEARLY THERE\nIn mid-October, Duffy the Disney Bear arrived in Disney California Adventure with a line of merchandise and a new photo op on Paradise Pier. ", "The bear is an import from Japan, where he is wildly popular and Disney is hoping he becomes a hit with American tourists like he did with the Japanese.", "\n\nSo far, lines have been short to meet Duffy and inside tips on his merchandise sales have reported varying degrees of success.", "\n\nHOTEL REMODELING\nOutside the parks, the Disneyland Hotel worked throughout the year on its major multi-year remodel project.", "\n\nIn mid-August Disney closed the wonderful Disneyland Hotel waterfalls forever. ", "The space is set to become a wedding lawn and nearby shops and restaurants are now gone and will be rebuilt as new dining and entrainment locations.", "\n\nA short walk from the park, HoJo Anaheim is conveniently located next to Disneyland with nearby dining and freeway access. ", "For more information on HoJo Anaheim, room rates, or to book your stay today, visit HoJoAnaheim.com/MiceChat.", "\n\nRooms now available for as low as $59 per night!", "\n\nThe HoJo Anaheim is now offering deeper discounts for Disneyland Resort Annual Passholders! ", "On select nights, Disneyland Resort Annual Passholders can get rooms for as low as $59 per night, plus tax. ", "For more information visit the HoJo Anaheim Innsider Boards which features an up-to-date listing of available nights, or call (714) 776-6120 and ask for In-House Reservations to book. ", "If your date isn't listed, the HoJo Anaheim still offers rooms for 20% off their best available rate if you call (714) 776-6120.", "\n\nExperience Disneyland at home with the HoJo Mattercam!", "\n\nSee Disneyland from home via a live rotating video feed of Disneyland Park and Disney California Adventure Park from the rooftop of Howard Johnson Plaza Hotel - Anaheim Resort! ", "The HoJo Anaheim's newly upgraded Disneyland webcam now features new views including video feeds of Disneyland's nightly fireworks! ", "Check it out now!", "\n\nCELEBRATE NEW YEAR'S EVE AT DISNEYLAND!The Disney Parks Blog released some planning info for this year's New Year's Eve festivities at the Disneyland Resort theme parks:\n\nNew Year’s Eve 2010 at Disneyland Resort\nThis year, there will be special entertainment added to the traditional holiday fireworks in Disneyland park and the spectacular new “World of Color” show in Disney California Adventure park.", "\nIn Disneyland park, hours are extended to 2 a.m., and stages will be set up around the park for countdown festivities. ", "DJs will encourage guests to dance into the new year at Sleeping Beauty Castle and small world Promenade while live bands will liven up Tomorrowland Terrace, Carnation Plaza Gardens and New Orleans Square, which will also be the site of three performances of “Fantasmic!” ", "at 9 and 10:15 p.m. and immediately following the midnight “Fantasy in the Sky” New Year’s fireworks show.", "\n\nAt Disney California Adventure, where park hours are extended to 1 a.m., there will be special entertainment along the Paradise Pier boardwalk and a “fountains and fireworks” countdown to midnight in Paradise Park, immediately followed by the first “World of Color” show of 2011. (", "Earlier New Year’s Eve performances of “World of Color” will take place at 8 and 9:15 p.m.) In Hollywood Pictures Backlot, revelers on the grid in “ElecTRONica” will heat up throughout the night and into the new year.", "\n\nBeginning at 6 p.m. New Year’s Eve, party hats and noisemakers will be available while supplies last in both Disneyland, at the Festival Arena near Big Thunder Ranch, and Disney California Adventure park, in Condor Flats near the vintage airplane just west of the entrance to Soarin’ Over California.", "\n\nROSE BOWL TEAMS TO APPEAR AT DISNEYLANDInfo on the Rose Bowl teams annual visit to Disneyland:\n\nThe theme park traditionally hosts the teams before the New Year’s Day game in Pasadena. ", "This year, players and coaches from Texas Christian University and the University of Wisconsin will hold press conferences before spending time in the park.", "\n\nDisneyland visitors may get a chance to see the players and coaches during a 2:15 p.m. ceremony the day after Christmas in front of Sleeping Beauty Castle, said Donna Fisk, a Disneyland Resort spokeswoman.", "\n\nWhile seats are reserved for Rose Bowl guests, visitors can find limited-viewing areas on the sides of the “hub” — the circular plaza and area in front of the castle.", "\n\nVisitors may get a glimpse of the teams, but Fisk warned that the day after Christmas is usually crowded at Disneyland.", "\n\nWhew, what a year! ", "It's been a pleasure being here every week with you as we explore the current goings-on at the Disneyland Resort. ", "I hope you all have a happy and safe New Year's Eve and please be sure to join me here again next week as we celebrate the new year festivities at Disneyland!", "\n\nWe're only here because of your support. ", "If you enjoy Dateline Disneyland every week please consider helping us out by donating a buck or two so we can pay the bills and keep the updates coming! ", "You can donate with a click of a button via PayPal — click here!", "\n\nAlso, remember that MiceChat has tons of new content all week long! ", "For even more Disneyland news, photos and information check out the In the Parks column on Fridays. ", "And MiceChat's Weekend Update, tours the world of Disney theme parks and other worldly sights and landmarks!", "\n\n|\nJoin us on Twitter and Facebook for great Disneyland news and information throughout the week!", "\n\nYou folks are amazing at getting this done weekly and with such quality (and GREAT Pictures). ", "For those of us who are thousands of miles away (Canada) we are eternally appreciative of your hard and dedicated work. ", "Such an amazing year behind us and you guys didn't miss a single paintjob! ", "Great work and THANK YOU!!!", "\n\nThank you for all the effort you've put in this year to keep us up to date on all the changes going on at the resort and thank you for all the effort that this year in review post must have taken.", "\n\nToday is the day my 2 week stay at the resort was supposed to start but due to various reasons I had to cancel my plans and bookings last month, so today I've been feeling a little down today, reading your year in review has put a smile on my face and I thank you for that Andy.", "\n\nBest of luck for the coming year to everyone, I look forward to reading your first post for 2011." ]

{ "pile_set_name": "Pile-CC" }

[ 0.024096385542168676, 0, 0, 0.008130081300813009, 0.011111111111111112, 0, 0, 0, 0, 0, 0.010752688172043012, 0.011363636363636364, 0, 0, 0, 0, 0, 0.025, 0.008849557522123894, 0.010752688172043012, 0, 0.011111111111111112, 0.009345794392523364, 0.017045454545454544, 0, 0.006134969325153374, 0.0070921985815602835, 0, 0.013513513513513514, 0.019230769230769232, 0, 0.004464285714285714, 0.008, 0.006802721088435374, 0, 0, 0, 0, 0.016042780748663103, 0.012448132780082987, 0, 0.007042253521126761, 0, 0, 0, 0.0070921985815602835, 0, 0, 0.006644518272425249, 0, 0, 0.011764705882352941, 0, 0.01020408163265306, 0.01079136690647482, 0.006688963210702341, 0, 0.00558659217877095, 0.015015015015015015, 0, 0, 0.011904761904761904, 0.017241379310344827, 0, 0, 0, 0.006578947368421052, 0.006578947368421052, 0, 0, 0.013513513513513514, 0.02631578947368421, 0.021739130434782608, 0, 0, 0, 0.011428571428571429, 0, 0, 0.007874015748031496, 0, 0.014285714285714285, 0.019417475728155338, 0, 0, 0.018633540372670808, 0.03260869565217391, 0.008547008547008548, 0.014705882352941176, 0.003861003861003861, 0, 0.01652892561983471, 0.00980392156862745, 0.01098901098901099, 0.014598540145985401, 0, 0.006711409395973154, 0.012096774193548387, 0, 0, 0.0084985835694051, 0, 0.013157894736842105, 0.008547008547008548, 0, 0.010101010101010102, 0.017857142857142856, 0.01694915254237288, 0.015151515151515152, 0.007352941176470588, 0.011904761904761904, 0.006578947368421052, 0.0078125, 0.007936507936507936, 0.024691358024691357, 0, 0.008, 0.009174311926605505, 0, 0, 0, 0.016304347826086956, 0.0078125, 0, 0.0111731843575419, 0.015151515151515152, 0, 0.0024691358024691358, 0, 0.007352941176470588, 0, 0.0035335689045936395, 0.004608294930875576, 0.009933774834437087, 0, 0.01282051282051282, 0.00966183574879227, 0.005952380952380952, 0.01652892561983471, 0, 0, 0, 0, 0.006493506493506494, 0, 0, 0.01, 0.027777777777777776, 0, 0.010416666666666666, 0, 0, 0, 0, 0.0035714285714285713, 0 ]

[ "CHECK OUT OUR SPORTS COLUMNISTS\n\nKain Colter has hundreds of college athletes in his cheering section, along with one proud uncle.", "\n\n“He’s smart and he’s not afraid to take a stand,” said Cleveland Colter on Tuesday.", "\n\nKain is the Northwestern quarterback, now graduated, who told a National Labor Relations Board in Chicago that college athletes are employees and should have the right to unionize.", "\n\nCleveland is the USC safety from the late 1980s who runs a catering company that provides food for Phoenix schools.", "\n\nHe also has trouble playing basketball, running or doing much of anything, thanks to a torn ACL during his junior year at USC.", "\n\n“And for some reason it hurts more in October than any other time,” he said, laughing. “", "Maybe that’s because that’s the month when it happened.", "\n\n“I had a partial tear and decided to let it heal on its own. ", "Then I got the operation in ’93. ", "In between I was in San Francisco’s camp and Houston’s, but I couldn’t pass the physical. ", "In my mind there was no way I wouldn’t be in the NFL. ", "In any event, things need to change.”", "\n\nColter did not get the operation because he couldn’t handle the thought of anesthesia. ", "When he was 3, his father died in a dentist’s chair because of a bad reaction.", "\n\nNevertheless, he thinks college players should access their own doctors and surgeons and rehab people, and he thinks the college athletic programs should pay for it.", "\n\nThat is only one of the issues pushed by Kain Colter and the CAPA (College Athletes Players Association).", "\n\nYou’re familiar with the basics.", "\n\nCollege players, particularly in football and basketball, are denied access to significant education. ", "Even at Northwestern, Colter said, he was dissuaded from taking chemistry as a freshman.", "\n\nNorth Carolina’s African and Afro-American Studies Department conducted several classes that never existed, but produced acceptable grades for its students, mostly football players. ", "The chairman of the department is under criminal indictment.", "\n\nA Sports Illustrated investigation of Oklahoma State revealed that players got cash for their good play but were kicked off the team and out of school for poor play or bad health.", "\n\nAnd throughout the sport, replica jerseys are sold without one penny trickling down to the players who wear them, an injustice not lost on Johnny Manziel or anybody else.", "\n\nA 2011 study conducted by CAPA and Drexel University calculated that Division I men’s basketball players and FBS football players who receive full scholarships had an average of $3,222 in out-of-pocket expenses.", "\n\nThat would not be a problem for the 17 FBS coaches who earned $3 million from their schools in 2013.", "\n\nNorthwestern paid its coach, Pat Fitzgerald, $2.2 million. ", "At least he is still coaching. ", "Mack Brown will get $2.7 million in 2014 for not coaching Texas, along with a $500,000 “special assistant” job.", "\n\nFootball is not seasonal anymore. ", "Colter said the Wildcats worked 40-50 hours a week during the season, up to 60 during August training camp.", "\n\nEven players’ tweets were subject to review, he said.", "\n\n“You have to sacrifice either football or academics,” Colter told the NLRB. “", "You’re not allowed to sacrifice football.”", "\n\nThe complaint is not specifically with Northwestern, though, and there is no talk of mass demonstration or wildcat, or Wildcat, strikes.", "\n\nThe reforms the players want are relatively simple and will not starve the fatted calf that is college athletics.", "\n\nAt issue is whether they are “employees.”", "\n\nCatherine Fisk, law professor at UC Irvine, observes that the NLRB has allowed students who work as medical aides to call themselves employees. ", "It hasn’t done the same with graduate students who teach classes.", "\n\nSince most football players will not go on to pursue professions in football, Fisk says one can “arguably” call them employees.", "\n\nHowever, there is a major kink in this process, one that makes it difficult to envision collective bargaining in Cleveland Colter’s or even Kain Colter’s lifetime.", "\n\nFisk points out that the NLRB serves Northwestern because it is a private university. ", "But public universities are governed by their state labor boards.", "\n\nThat would be a problem in Southern states that love football and disdain unions with equal passion.", "\n\nIn Alabama and North Carolina, collective bargaining between the government and its workers is prohibited. ", "Other Southern states allow bargaining with very limited groups.", "\n\nIt’s difficult to see Alabama legislators welcoming football players to the bargaining table.", "\n\nNot unless private schools, or schools in more labor-friendly states such as California, use better working conditions as an extra benefit in recruiting.", "\n\nAnd perhaps that advantage would be the thing that convinces the NCAA to acknowledge the dawn of the 21st century and sit down with CAPA.", "\n\nIt never would consider doing so unless Kain Colter, like his uncle, had braved the risk of standing up.", "\n\nUser Agreement\n\nKeep it civil and stay on topic. ", "No profanity, vulgarity, racial\nslurs or personal attacks. ", "People who harass others or joke about\ntragedies will be blocked. ", "By posting your comment, you agree to\nallow Orange County Register Communications, Inc. the right to\nrepublish your name and comment in additional Register publications\nwithout any notification or payment." ]

{ "pile_set_name": "Pile-CC" }

[ 0, 0.011764705882352941, 0.005494505494505495, 0.008547008547008548, 0.0078125, 0, 0, 0, 0, 0, 0.018518518518518517, 0, 0, 0, 0, 0.009345794392523364, 0, 0, 0.011363636363636364, 0.005434782608695652, 0, 0.0055248618784530384, 0.005813953488372093, 0.014084507042253521, 0, 0.01639344262295082, 0, 0.009009009009009009, 0, 0.009345794392523364, 0, 0, 0, 0.007246376811594203, 0, 0, 0.0136986301369863, 0, 0.007751937984496124, 0.012121212121212121, 0.011363636363636364, 0, 0, 0, 0, 0, 0, 0.014388489208633094, 0.009433962264150943, 0, 0, 0, 0.00975609756097561 ]