Datasets:

CleverShovel

/

paper_reviews

like 1

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: DOES SILVOAGROPECUARY LANDSCAPE FRAGMENTATION AFFECT THE GENETIC DIVERSITY OF THE SIGMODONTINE RODENT OLIGORYZOMYS LONGICAUDATUS? Review round: 2 Reviewer: 1

Basic reporting: The manuscript improved significantly its quality from the changes made. The incorporation suggested in figure 1 of the metric scale as well as the palette of colors and the image in google earth undoubtedly facilitates the identification of the different environments and their interpretation.Likewise, the division into sections of data analysis orders its reading and facilitates its interpretation satisfactorily Experimental design: The modifications made improve the interpretation of the results. The methodology was described in greater detail facilitating greater understanding and interpretation of the analyzes. Validity of the findings: In general terms, the explanations included in this new version are consistent with the conclusions reached by the authors. The results have been discussed in greater depth and also, the included bibliography contributes satisfactorily to improve the quality of the manuscript. However, since the data were taken in a given period of the year it is necessary to be discussed within that context. Therefore it should be clarified in the discussion that the interpretation that is being made corresponds to the period of less mobility of O. lonficaudatus since in the reproductive season similar results could not necessarily be observed. Additional comments: No comment.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: DOES SILVOAGROPECUARY LANDSCAPE FRAGMENTATION AFFECT THE GENETIC DIVERSITY OF THE SIGMODONTINE RODENT OLIGORYZOMYS LONGICAUDATUS? Review round: 2 Reviewer: 2

Basic reporting: This paper reports a landscape genetic study of the sigmodontine Oligoryzomis longicaudatus (“Colilargo”) based on the analysis of 5 populations sampled in 5 patches located in a fragmented landscape with different types of soil use. The aim of this study is to evaluate the effects of landscape fragmentation on genetic diversity, and to investigate the most probable landscape characteristics of commuting “least-cost corridors” among patches The research question is well defined. Experimental design: Although a single marker was included in the analysis, it shows enough variability as to reveal patterns of distribution of variability in the landscape. Geneland is not the best approach for this analysis. Authors would have to justify their inclusion, describing that mitochondrial sequences were treated as a single loci, with many allelles (i.e. haplotypes). Validity of the findings: The fact that the studied species is a reservoir of the Hantavirus emphasizes the importance of the results for regional sanitary controls, in order to reduce the risk of humans to contract HPS Additional comments: All suggestions made by the reviewers were properly adressed. Therefore, I have no further comments.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROSPECTIVE STUDY OF THE MOTIVATIONAL AND HEALTH DYNAMICS OF INTERNET GAMING DISORDER Review round: 1 Reviewer: 1

Basic reporting: The present study recruited 5,777 American adults and applied self-determination theory to examine how motivational factors influence, and are influenced by, IGD and health across a six month period. The sample and the longitudinal design is very impressive. Overall, it is a very important article for this filed. Experimental design: Research question well defined, relevant & meaningful. The authors should take "internet addiction test" into their consideration for the examination for the participants if they collected this score. Validity of the findings: Data is robust, statistically sound, & controlled. Additional comments: The present study recruited 5,777 American adults and applied self-determination theory to examine how motivational factors influence, and are influenced by, IGD and health across a six month period. The sample and the longitudinal design is very impressive. Methods described with sufficient detail & information. The results were robust and interesting. Overall, it is a very important article for this filed.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROSPECTIVE STUDY OF THE MOTIVATIONAL AND HEALTH DYNAMICS OF INTERNET GAMING DISORDER Review round: 1 Reviewer: 2

Basic reporting: The current study offers a substantial advance in our knowledge and understanding of a complex and controversial disorder. Furthermore, the authors should be commended for taking an open science approach to the study, and pre-registering their analyses. In sum, I believe that this paper has a vitally important contribution to make to the IGD literature – not just in terms of the findings, but also as a model example of the way psychological experiments should be conducted. Experimental design: The authors present here a large-scale study looking at the presence of Internet Gaming Disorder (IGD) symptoms in a substantial sample of American participants, and apply self-determination theory to assess if and how presence of IGD affects health and behavioural outcomes over a six month period. The research is robust, the research questions are well-defined, and fill an important gap in the research literature. Validity of the findings: The study presents some fascinating and novel findings. IGD appears to be disruptive of some health factors in terms of psychological need supports, but the authors present evidence of no direct effect of dysregulated gaming on health over time. While IGD symptoms appear to be moderately stable over the six month period, the authors in fact show evidence of a drop in problematic symptoms over that time – no one who met a diagnostic threshold at the start did so at the end. This finding has important consequences for the theoretical framework within which we consider IGD – currently, there is some debate as to whether it should be framed as either a substance abuse disorder, or as more of an impulse control disorder. Before we can make any headway in developing effective ways of assessing and treating IGD, this debate first needs to be settled. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROSPECTIVE STUDY OF THE MOTIVATIONAL AND HEALTH DYNAMICS OF INTERNET GAMING DISORDER Review round: 1 Reviewer: 3

Basic reporting: The most important concern is the clarity of presentation of the many analyses. There are only two tables and two figures for the many analyses presented, and statistical estimates are presented in only one of these. Path analyses and SEM models are often reported with statistical estimates included. If possible, reporting the various path analyses in clearly-labeled and described figures would be useful. The reporting of fit statistics would be useful as well. Although the main model is described as a fully saturated model, the wording of that sentence (335) may make it appear as if this is a result the fit of the model rather than a property of saturated models by definition. This line would benefit from being reworded. Further discussion of modeling is covered in the Validity section. Second, there are some instances of terminology that were unfamiliar or required some thought/research to understand. While psychological need satisfaction is well-known. I was unable to find many references to "need supports". For example, it appears in work by Deci and Ryan, but not in the works cited in the paper (around line 91). In addition, use of the term "need supports" in a sentence is a little confusing because the word "supports" is more often used as a verb. For these reasons, I would suggest the authors consider using the term “need satisfaction” or even reword sentences containing “need supports” to improve clarity. Likewise, “motivational dynamics” is another term that may not be easily understood outside this field. While these terms are certainly relevant and accepted within certain areas of study, the paper would be better understood by those outside of these areas if other language is used. Experimental design: This paper uses a design common to survey studies. Any concerns are more related to reporting and validity and are covered in those sections. Validity of the findings: One of the major strengths of this paper is the potential representativeness of this sample. However, external validity is a concern. Some properties of the sample make it seem less representative than expected. The mean age is 46, which seems high for an Internet gaming sample. Without performing calculations, it is difficult to tell the racial makeup of the sample, which is another important factor in determining representativeness. There is no information on socioeconomic status. The sample has a higher proportion of females than males; is that in line with expectations for a sample of Internet gamers? In understanding how the important results of this study may relate to other populations of Internet gamers (e.g., younger, mostly male), it is important to have a little more explanation about the intent, recruitment and data collection in the original study and a critique of how the study sample may differ. A final note about the study procedure: the method of informed consent should be specified. What does it mean that there was a double opt-in? The time period of the analysis is also a consideration, especially as it relates to the sample’s age. Would a 6-month period be adequate to capture causal effects, especially if the disorder itself is short-term in nature? Some mention of this potential problem would help contextualize the findings. The reporting of analyses also brings up questions of validity. As stated above, the saturated model is described as having perfect fit, but this tells us nothing about the meaning or potential invalidity of the estimates as it is a property of a saturated model. Several methods can be used to evaluate the appropriateness of such models (see a standard text on structural equation modeling such as (Kline, 2010). If I understand correctly from the code, a multigroup analysis by sex was conducted. This would be an example of an alternative model that could be tested and reported. Another concern is how the use of FIML was conducted. More explanation of the steps involved in lavaan would be useful for those who are less familiar with this package. For example, the authors could specify, “The full information maximum likelihood estimator was used in the lavaan package in R by …” A final concern major consideration is the wording of one of the IGD indicators. Starting on line 255, the paper discusses the “distress” criterion of IGD. In fact, the wording of the IGD criterion is “clinically significant impairment or distress”. The self-report of distress may differ from the self-report of impairment, and the importance of this clinical judgment is, of course, a debate in the field. I would be especially interested to know if self-report of distress differed by sex or age group, for example. Additional comments: This paper reports a very thorough analysis of representative longitudinal data on psychological needs, Internet gaming disorder and health using a well-researched and thoroughly developed rationale. The strengths of the study are its clearly-defined hypotheses, its transparency, and the connection of hypotheses and results. As the Introduction states the intent to bridge the gap in the literature between self-regulation and addiction, this reviewer reports from the perspective of a researcher trained in public health and addictions with suggestions for reporting that will make the strengths of the article more accessible to those in the biomedical sciences. Overall, the authors have conducted a well-thought out study that would benefit from some improvements in reporting to clarify limitations related to sample and to the complex analyses. I have outlined other concerns or provided more information about the items summarized above in a line-by-line analysis below. It is certainly not necessary to respond to each line, however. I welcome the chance to be involved in reviewing a revised version of this manuscript. Line by line review Line 64: Individual indicators in line with APA guidance- The word indicators has a specific statistical meaning. It would be better to use criteria (as they are referred to in the DSM) or symptoms. The word appears again in line 245 and other places. Line 76: The reference provided refers to a very specific population, and thus may be overstated. Since there is also good support for the opposite (i.e., detrimental effects on populations) ,it would be better to find additional support for that conclusion (e.g.,(Hofferth & Moon, 2012)) and to critique studies that find that time gaming is associated with harm at the population level (e.g., (Faulkner, Irving, Adlaf, & Turner, 2014; Festl, Scharkow, & Quandt, 2013). If the concern is a lack of longitudinal studies that provide additional support for causal determination, that also has conflicting evidence in the literature (e.g., (Gentile et al., 2011). With regard to the Scharkow et al. study in 2014, it’s also noteworthy that no longitudinal associations between PG and well being were found for adolescents. Lines 78-82: These lines seem to be making the case that PG is caused by dysregulation, but the statement on line 82 is that it is “dysregulating”, implying that it causes dysregulation. Line 91: I was initially confused by the term psychological need supports, as I have never seen this term in the biomedical literature. I did make a point of reviewing some of the cited articles, and I see that they use other terminology such as need satisfaction. I would recommend that if this article is meant to bridge the literature, terminology that is more straightforward would be easier to understand. Line 101 is another example of where the terminology makes the sentence a little confusing. Some alternatives might be “need satisfaction”, which seems more common in the SDT literature, or even rewording to remove jargon. For example, line 99-100 could be worded as “Having psychological needs satisfied is associated with better mental, physical and psychosocial health.” 114: A similar problem is reflected here, where the claim that Psychological Needs Reduce Dysregulation is confusing. I think it is having psychological needs met that reduces dysregulation. 144-147: In discussing the displacement of activities, this statement seems out of place. 155: Obsessive passion for video games? Internet games? Activities in general? 157: Gaming and Everyday Activities. While the previous section uses an SDT framework to discuss the potential mechanisms of effects of engagement on psychological needs, the section wants a framework. The authors might consider mentioning the displacement hypothesis (Valkenburg & Peter, 2007). Another concern with this section is that it discusses the importance of social activities without providing a reason that reduced social activities might be associated with health. Social participation is a basic function of everyday life, and theories such as the stress buffering hypothesis (Cohen & Hoberman, 1983) might help support the importance of social activities for health. 175: Motivational dynamics is another term that is somewhat unclear at first. 220: From a public health perspective, it would be important to have more information about the sample and study. For instance, how was the survey conducted? Was it conducted online, by telephone or in person? How did the sample of Internet gamers compare to the overall sample? I notice that the average age is 46.2; how does this compare with studies using other population samples in this topic area? The reporting of race is concerning, as it is difficult to tell the representativeness without performing the necessary calculations to determine what percentage of the sample is White (2234/3146). Percentages should be included here. 235-243: Was informed consent provided, or was the double opt-in considered informed consent in some way? 255: The wording of the proposed IGD definition is “ impairment or distress”. Obviously the impairment piece is another point of contention in the IGD debate, especially as the reliability of respondents to truthfully report impairment is called into question for any mental disorder. The paper should address the discrepancy between the wording for this survey item and the requirement for “clinically significant impairment or distress” in the proposed IGD diagnosis. 279: Please include a sample question for those who are new to these scales. 328: I’m confused as to how the inclusion of T1 IGD scores contributes to FIML estimation when these variables are a part of the specified model. Perhaps the sentence needs to be reworded to describe which other variables, if any, were also included that were associated with attrition. 335: Reporting of model fit. The results section describe the model as saturated and producing perfect fit with d.f.=0. Since saturated models by definition produce perfect fit, it is necessary to use other methods to assess whether the fit is in line with expectations. The authors might consider conducting hypothesis testing comparing their chosen model with other potential models, for example. 353: Small significant betas may result from the large sample size. It is good that they were consistent with the author’s predictions in terms of being in the appropriate direction, but what is the theoretical implication of a beta of magnitude 0.06? Expressing standardized betas in the form of change in expected outcome for standard deviation of the chosen indicator would make this more clear. (e.g., for every standard deviation change in X, Y was expected to increase by ___) It also seems that at times that unstandardized betas are reported (e.g., Table 2). This would be good to clarify with each set of models discussed, or report one or the other consistently. When development of a construct is the question, unstandardized estimates, including residual covariance, allow for a better understanding of heterogeneity in change and the ability of the model to explain relationships. This is another potential consideration. 390: I’m not sure, but it seems that “need supports correlates” might be a typo. 409: It doesn’t seem to me that 10+ years of the study of problematic online gaming (e.g (Kraut & Seay, 2007)), would qualify for being in the early stages. That’s not to say that the current state of the evidence is enough to correctly conceptualize a formal disorder. Nonetheless, I think it’s an overstatement to say that work is in an early stage. Beginning on line 415: This paragraph has very important implications for IGD and similar diagnoses. The finding that a diagnosis of IGD based on satisfying a number of criteria does not hold at 6 months implies that the 12-month time criterion may lead to brief periods of “disorder” being missed. Have there been other suggestions as to a relevant time period for IGD? This would be important to know and discuss. Also, the paper describes options for time periods of addictive disorders are chronic or episodic. From a more medical perspective, addictions are often seen as an “absorbent state” (as are other mental disorders) where individuals never truly recover, but suffer from occasional episodes. Problems related to gaming might also represent a phase, however, which is more in line with the maladaptive coping hypothesis (Kardefelt Winther, 2014). In this case, rather than chronic or episodic, quickly resolving might be another possible way to describe problems. The potentially brief nature of disordered gaming also has implications for the 6-month lag in the current study. The authors are encouraged to discuss the implications for their finding in terms of the potential ephemeral nature of online gaming-related problems, how their sample may have influenced this finding in their analysis, and also how these may differ by populations. 470: I was not familiar with the word choiceful and had to look it up. I see it is British English, and while this is of course perfectly acceptable, I am still in favor of rewording to make it even more clear. 483: The paper states that IGD symptoms are linked to health “partly because they undermine need supports”, yet previous it seems that IGD symptoms were only indirectly linked to health. Some way to organize results more clearly (additional tables with clear description of models tested, e.g.) would be useful. 498-501: The literature cited here refers to samples of adolescents, yet the sample in the study is of middle-aged adults. The authors should consider how mechanisms may differ in their sample. 555-556: “absence of supportive environments” has a much different connotation in the public health literature, speaking mainly to child development and a life course approach. My understanding is that motivational dynamics as studied here may the result of environments experienced during upbringing, but are not necessarily limited to this. That phrase may benefit from being reworded. Figures: The results of the hypothesized model (the betas) would be much easier to understand if they were indicated in the path model rather than solely in the text. Figure 1 note: It is unclear which nested models were tested and how covariances were evaluated.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROSPECTIVE STUDY OF THE MOTIVATIONAL AND HEALTH DYNAMICS OF INTERNET GAMING DISORDER Review round: 1 Reviewer: 4

Basic reporting: This manuscript describes a study of the motivational and health dynamics of Internet Gaming Disorder. The task was well controlled. The study applied self-determination theory to examine the relationship of IGD, motivational factors and health across a six month period. The results confirmed that IGD indicators are stable and that they and basic psychological need supports have a reciprocal relationship over time. The need supports promoted health and served as a protective factor against IGD. This is an important topic and the question is of wide interest for researchers working in IGD. In discussion relate to Internet addiction management, I will recommend to include 10 Hz rTMS can be used to treat addiction. Reference (1): Shen et al. 10-Hz Repetitive Transcranial Magnetic Stimulation of the Left Dorsolateral Prefrontal Cortex Reduces Heroin Cue Craving in Long-Term Addicts. Biol Psychiatry. 2016 Aug 1;80(3):e13-4. (2) Transcranial magnetic stimulation of dorsolateral prefrontal cortex reduces cocaine use: A pilot study. Eur Neuropsychopharmacol. 2016 Jan;26(1):37-44. In addiction, tDCS helps to treat addiction. Reference: (1) Meng et al. Transcranial direct current stimulation of the frontal-parietal-temporal area attenuates smoking behavior. J Psychiatr Res. 2014 Jul;54:19-25 (2) Wang et al. Transcranial direct current stimulation of the frontal-parietal-temporal area attenuates cue-induced craving for heroin. J Psychiatr Res. 2016 Aug;79:1-3. Experimental design: This study has clear description in the material and methods. The work conduct in conformity with the ethical standards of the field. Validity of the findings: The statistical methods are used valid in this study. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE RELATIONSHIP BETWEEN SUSTAINED ATTENTION AND AEROBIC FITNESS IN A GROUP OF YOUNG ADULTS Review round: 1 Reviewer: 1

Basic reporting: The present manuscript is relatively well write. However, several part need to be carefully reread to correct several inattention mistakes. - Line 51: incongruent date and time - Line 69-71: the sentence begin in english and finish in spanish - Line 79: comments in spanish and annotation References in the text should be check. There are several redundancy in name or bracket wrongly used line 73, 79, 317. I am not relevant to judge the quality of English. The working hypothesis could be detailed at the end of the introduction. The table 1 is cited at page 7 but only insert in page 11. Experimental design: The present study aimed to investigate the impact of aeobic fitness on sustained attention in young adults. A high-fit group and a low-fit group completed a 5' PVT task (two times, respectivelly before and after an oddball task) and a 27' oddball task. Method: Line 159: precise that the cognitve tasks were performed at rest. Line 160: precise taht the incremental test was performed in order to to determine the aerobic fitness caracteristic of the subjects (i.e., VAT, VO2 at VAT, power at VAT). The description of the VAT determination test should be presented in a separate paragraph like the 2 cognitive tasks, intitulated "VAT determination test" for exemple. The sentence "The filled circumference was presented for 500 ms" make the description of the PVT confusing, I suggest to delete it or move it somewhere else in the paragraph (line 185 for example). Line 198: Contrary to the PVT, you encourage your subjects to respond accurately, not quickly. Why? this could explained part of your results, especially the afct that you have no effect on RT in the oddball task. This need to be discuss. Line 202: The authors indicate to late (line 208-209) that the oddball task contained 2 distinct blocks. This information shuld be given earlier (line 202). Did the subjects have a short break between the 2 blocks? Validity of the findings: The results of the oddball task would be more clear if the results would be presented in two separated paragraphs. The first one concerning 1/ Hit trials and the second 2/the commission errors. Line 238: precise that the interaction you are talking about is the fitness x TOT interaction. Figure 1 and 2: please add symbols representing the significant effects on the graph. Change the title of the y-axes, respectively by "Traget trial accuracy" for the Fig 1 and "Non target trial accuracy" for the fig2. Add the means in the text in the results section. Line 256-258 should be move in the discussion section or deleted. The results confirm that high-fit subjects performed better on the PVT task than low-fit subjects. However, no effect of TOT task and no interaction with aerobic fitness was observed. We could have expected a degradation of the perfomance over time (between pre and post PVT), at least for low-fit subjects. How do you explained that? Line 295-297: reverse the sentence. You should discuss the fact that the oddball task is perhaps not enough sensitive to highlight the effect of fitness on sustained attention. Moreover, the accuracy instructions given for this task in particular could have altered the results. Line 315: change VO2 by VO2 at VAT. Do you have any idea of a better predictor? Additional comments: The present study is of interest, however some modifications are necessary before publication. I hope that this comments will be helpful.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE RELATIONSHIP BETWEEN SUSTAINED ATTENTION AND AEROBIC FITNESS IN A GROUP OF YOUNG ADULTS Review round: 1 Reviewer: 2

Basic reporting: This current study examined the relationship between sustained attention and aerobic fitness in young adults using oddball and PVT cognitive tasks in young adults. Results indicated that higher-fit adults maintained response accuracy over time, while low-fit adults suffered a decline of response accuracy throughout the task, suggesting that higher fit adults have a great sustained attention. This manuscript was well written and the results were significant. Minor change: 1. Line 70-71: please rewrite the sentence as it was not written in English. 2. Line 155: "approx" is not a complete word. 3. Line 210: "using 1-way between-groups design" seems not clear regarding of the proper description for statistical analysis. Experimental design: no comments Validity of the findings: no comments. Additional comments: no comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE RELATIONSHIP BETWEEN SUSTAINED ATTENTION AND AEROBIC FITNESS IN A GROUP OF YOUNG ADULTS Review round: 1 Reviewer: 3

Basic reporting: This manuscript focuses on the relationship between aerobic fitness and sustained attention assessed by a PVT and an oddball task emphasizing response time and accuracy, respectively. While I believe that this manuscript is interesting, there are a number of potential issues that should be addressed prior to this manuscript being considered for publication.  Please make sure the format of in-text citation and reference is consistent throughout the manuscript as well as in line with PeerJ’s guideline. For example, be consistent in terms of whether capitalizing the first letter of each word in article and journal title in the reference section. Also, check in-text citation such as line 79 and 317.  Some texts in the manuscript were not written in English (line 70, 79..). Please proofread and modify accordingly.  I would recommend the authors to change the title given that the current title reads like only oddball task was used.  Line 81: the authors may need to explain what were the neuroelectric indices for sustained attention or response preparation in Luque-Casado et al. (2016).  Line 309: these two citations regarding brain activation during exercise are odd at here. Instead, the authors should cite prior research examining the effects of fitness or chronic participation in physical activity on structural and function adaptations in the brain. Minor issues 1. Some grammar errors need to be corrected such as line 6 in the abstract ("divided in" should be "divided into") and line 313 ("limit" should be "limits"). Please proofread or have a English native speaking colleague review the manuscript. 2. Line 46: I would recommend the authors avoid stating “any type of information processing” as there may be processes that require little of sustained attention such as some bottom-up or automated processing. 3. Line 51: there are irrelevant numbers at the end of the sentence. 4. Line 197: it may be helpful to add “non-target” before “long line”. Experimental design:  What is the unique contribution of this study? Why is it important to examine the relationship between aerobic fitness and sustained attention using an oddball task emphasizing accuracy? In other words, does response accuracy during an oddball task reflect one aspect of sustained attention that is different from RT during PVT? The authors may need to discuss why investigating the effect of fitness on accuracy sustained attention task is necessary for filling the knowledge gap in the field.  Line 124: It is not clear why the authors adopted two blocks of 5 min PVT task separated by a 27 min oddball task. If the purpose of this study was to replicate Lique-Casado et al., (2013, 2016), why not used a 10-min block or a 60-min block? Please justify.  Line 131: please report the effect size that was powered to be detected based on the sample size in the current study.  Line 134: Were participants allocated to low- and high-fit groups according to the frequency of weekly training? Does the training mean sports training or physical activity? And how to define training and physical activity in terms of intensity? The description (line 134-138) may need to be modified for better clarity. Minor issues 1. Line 208: The authors may consider moving this sentence to the Oddball task section. Validity of the findings:  Line 260: The significant fitness main effect on non-target trials does not support the selective improvement hypothesis discussed in line 302 given that high-fit participants outperformed their low-fit counterpart for both target and non-target trial types. The authors need to include trial type as a factor to test whether the beneficial effect of aerobic fitness on target trials is larger than that of non-target trials.  The data from PVT did not show an impaired performance over time for both groups. Was PVT used in this study a valid measure of sustained attention?  Line 321-322: Instead of their own belief, the authors should discuss what the existing literature found. For example, McMorris et al. (2016) and Voss et al. (2013) provided very comprehensive reviews regarding the possible mechanisms to account for the beneficial effects of chronic and acute physical activity on cognition.  Line 289: the statement “this later findings supports the PVT as a suitable tool to measure group differences in terms of RT performance in a vigilance task involving high temporal uncertainty of target onset” is fine. However, such fitness main effect may be not due to modulation of sustained attention because RT was maintained over time in both low- and high-fit groups. The authors need to explain what is contributing to this fitness main effect on PVT performance and why it is not consistent with prior studies.  It is not clear why the authors performed a correlational analysis among measures of aerobic fitness and sustained attention. Such analytical approach should generate similar results derived from factorial analyses in a cross-sectional study. It is strange that a fitness main effect was observed using factorial analyses but the correlation between fitness and measures of sustained attention is null. The related issues are below I. The result of the correlational analysis is problematic because it suggests that aerobic fitness has no impact on sustained attention, which is contrary to the results derived from the factorial analysis. If the authors intended to include both of these analyses, they may need to explain such discrepancy. II. It is not clear if aerobic fitness was correlated with the overall accuracy across block 1 and block 2 of the oddball task or with accuracy for each block separately. III. Line 315-317: the two points based on the null correlation are confusing. In this study, VO2 reflected aerobic fitness. The null correlation only suggests that VO2 is not a good predictor for sustained attention. Similarly, the null correlation suggests that aerobic fitness, which is essentially VO2, cannot explain sustained attention capacities. Regardless, the correlational analysis performed in this study cannot generate any conclusion related to the so-called cardiovascular hypothesis. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE RELATIONSHIP BETWEEN SUSTAINED ATTENTION AND AEROBIC FITNESS IN A GROUP OF YOUNG ADULTS Review round: 2 Reviewer: 1

Basic reporting: no comment Experimental design: no comment Validity of the findings: no comment Additional comments: The authors have responded to all my comments. The manuscript has been greatly improved and deserves publication.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE RELATIONSHIP BETWEEN SUSTAINED ATTENTION AND AEROBIC FITNESS IN A GROUP OF YOUNG ADULTS Review round: 2 Reviewer: 2

Basic reporting: 1. Line 133: It may be helpful if the authors could cite studies that tested the reliability of short PVT in adults. If no such evidence exists, the authors may consider replacing “infants” by other terms that indicate children aged ~11 yr according to Wilson et al. (2010) 2. To me, “mind reset” is a perfectly fine explanation for the lack of session effect in the PVT. However, another possibility is that the PVT (two 5’ blocks separated by oddball task) employed in the current manuscript was simply not demanding sustained attention/vigilance enough to induce the RT increment from the first to second block. If I understand correctly, line 284-287 might be intended to suggest that it is possible that PVT is a valid measure of sustained attention/vigilance even when time-on-task effect is not apparent or not observable. However, it may be helpful if the authors could explain more about this idea as well as the cited studies (e.g., Dorrian et al., 2005; Drummond et al., 2005; Lim & Dinges, 2008) with better clarity. The authors then stated that “High-fit and low-fit participants appear to differ in their overall RT performance, but not in the effect of session. This would seem to nuance our statement of the superior vigilance capacities of high-fit participants.” I am not sure if the first sentence is based on prior studies or the current manuscript. If it’s referring to prior studies, please provide citations. If the sentence is based on the current study, it is redundant because the following sentence (Line 290-291) repeats the similar thing. Thus, I would recommend revising the second half of this paragraph (line284-294) for better readability 3. Please check the use of tense throughout the manuscript. For example, Line 101 Participants “were” asked; Line 227: Probability values “were” reported; Line 228: ANOVA “were” reported… Experimental design: NA Validity of the findings: NA Additional comments: I commend for the authors' effort on addressing all of my concerns. This manuscript is substantially improved and can contribute to the literature. I believe this manuscript can be accepted for publication after the minor revisions listed above.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: TRICHOPLAX ADHAERENS REVEALS A NETWORK OF NUCLEAR RECEPTORS SENSITIVE TO 9-CIS-RETINOIC ACID AT THE BASE OF METAZOAN EVOLUTION Review round: 1 Reviewer: 1

Basic reporting: Review on the manuscript by Novotný et al. “Trichoplax adhaerens reveals an endocrine-like network sensitive to 9-cis-retinoic acid at the base of metazoan evolution” In Srivastava et al. 2008, four different nuclear receptor (NR) family transcription factors were reported to be encoded in the genome of Trichoplax adhaerens, a basal animal that is morphologically very simple. The four Trichoplax family members, RXR (NR2B), HNF4 (NR2A), COUP (NR2F) and ERR (NR3B), show a high degree of similarity to their human homologs. NRs are normally defined as ligand-activated transcription factors that allow the regulation of target genes by small lipophilic molecules such as hormones (thyroid hormones, steroids), morphogens (retinoic acid) or dietary components (fatty acids). In the present article, Novotný et al. explore the role of NRs in Trichoplax. First, they show that one of the four NRs of Trichoplax, the retinoid X receptor homolog (TaRXR), is highly conserved (about 66% identity to vertebrate homologs) and is able, like the human homolog, to bind to 9-cis-retinoic acid (9-cis-RA) in the low nanomolar range. To show this, they have used an assay with radioactive 9-cis-RA established earlier (e.g. used to describe the affinity of the RXR of the cnidarian Tripedalia cystophora by Kostrouch et al. 1998). Novotný et al. also show that 9-cis-RA increases the expression of a malic enzyme they had identified in Trichoplax. They suggest that this could reflect a natural regulation of this metabolic enzyme by activated TaRXR. Moreover, the authors observed a change of the expression pattern of three of the four NRs induced by 9-cis-RA. Apparently, the authors also observed a difference in the expression patterns of NRs when they compared animals of different sizes. In the last part of the manuscript, the authors report that the food source has an impact on the size and growth rate of Trichoplax. I have to admit that I did not quite understand why these observations have been added to the manuscript about NRs and the effect of 9-cis-RA on TaRXR. The observations on the effect of food on Trichoplax are not really connected with the rest of the report. Somehow, when I read the abstract, I thought that addition of 9-cis-RA to the culture would have changed the shape, size or behaviour of the animals – it almost goes without saying that this would be very interesting. Have such experiments been performed and no changes were observed? Do the authors think that components of the algae might be ligands for the NRs? Another possibility to link the two sections would have been to test for changes of the expression level of different NRs upon feeding the animal with different algae. Overall, I find the observations presented in the manuscript interesting, but several aspects were not entirely clear to me. I therefore suggest that the authors re-analyze their data and carry out additional experiments before considering publication of their findings. Moreover, I do not agree with the authors that the results indicate the presence of an “endocrine-like network” in Trichoplax. Please note that the current title of the manuscript does not really make sense. The term “endocrine-like network probably refers to the possible network of NRs found in the study, but the term endocrine appears a bit far-fetched as it implies the presence of an endocrine system in the animal, for which the present study does not provide clear evidence. Major points that need to be revised: 1) In the Introduction, the authors do not make sufficiently clear why they have studied the NRs of Trichoplax adhaerens in particular. For example, they write “that Life on Earth began 4.1 to 3.5 billion years ago (Bell et al. 2015) with the appearance of the first unicellular prokaryotic organisms that subsequently evolved, in part, to multicellular lifeforms forming the kingdom Metazoa that have specialized tissues for digestion, regulation of homeostasis, locomotion, perception, analysis of the environment and reproduction.” This is obviously not correct: animals are eukaryotic organisms. Several eukaryotic lineages are multicellular: animals, fungi, plants and others. Here it would be advisable to shortly explain what we know about the evolution of NRs and their ligands in animals (and other eukaryotes), for example as has been done in Kostrouchova & Kostrouch 2015, for example. Trichoplax with its relatively simple set of NRs represents an animal at the base of the animal kingdom and could therefore shed light on an early set of NRs in animals, if that was the aim of the study. I think that this point is currently more clearly outlined in the Discussion, a section of the manuscript written in a clearer fashion than the rest of the manuscript in my opinion. 2) Remarkably, Trichoplax has a set of NRs that are highly similar to human NRs, in particular TaRXR is highly conserved. The conservation is shown by a sequence alignment, but I did not understand why only few organisms are represented in the alignment (and in the tree). Why, for example, are model cnidarians like Nematostella or Hydra not included? Don’t they have the particular receptor? What about sponges and ctenophores? What about bilaterians? This should be explained and the data set, if necessary, expanded. I suggest to improve the representation of the sequence alignment by clearly indicating the domains (DBD & LBD) and residues involved in 9-cis-RA. If these residues were indicated, the few non-conserved residues do not need to be outlined in the text (line 294). Check for example Gutierrez-Mazariegos et al.. 2016. Evolutionary diversification of retinoic acid receptor ligand-binding pocket structure by molecular tinkering. R. Soc. open sci. 3:150484 (PMID: 27069642). It also seems that the authors are unaware of a study, in which the set of Trichoplax NRs was investigated already (Baker ME. 2008. Trichoplax, the simplest known animal, contains an estrogen-related receptor but no estrogen receptor: Implications for estrogen receptor evolution. Biochem Biophys Res Commun 375:623–627. PMID: 18722350). By the way, I did not understand in the manuscript (line 280) whether the known TaRXR sequence (Srivastava et al. 2008) was used or whether the authors generated a revised/improved sequence of the TaRXR. If so, what exactly was improved? Furthermore, I suggest to omit the phylogenetic analysis. As far as I understood, the important point was to show that the 9-cis-RA binding site is conserved between man and Trichoplax. Probably there is a protein structure that can be used to discuss the conservation and binding site. A tree that is based only on a few sequences does not provide much information and might be biased, because only selected sequences have been used. I also suggest to omit Table 1, because a sequence comparison based on a few Blast search results does not provide much information. For both data sets, a more rigorous analysis would be necessary. Again, the authors should compare their results to the above mentioned study by Baker ME. 2008. 3) In the section about 9-cis-RA binding, the authors only state that 9-cis-RA binds with high affinity and specificity, but do not explain in the text nor in the figure legend how the experiments have been performed. Currently, the reader has to go to the Method section to find out that this was done using radioactive 9-cis-RA (how was radioactivity measured?) which was added to recombinant proteins expressed in E. coli. Why isn’t the purity of the proteins shown by SDS-PAGE, for example? Was TaRXR used as a GST fusion protein or was the cleaved protein used in the experiments shown? How does the affinity measured compare to the affinity of 9-cis-RA to the human (or others) protein reported in earlier studies? That “this was clearly observed in repeated experiments” (line 319) should not be stated in the Results, rather the figure legend should make clear that the experiments were carried out several (how often?) times, also by adding error bars. 4) The upregulation of the malic enzyme in Trichoplax has been demonstrated by quantitative PCR (Fig. 4). However, no control of a housekeeping gene like actin or another enzyme has been performed to make sure that really similar amounts have been used. This also concerns the experiments shown in Fig. 5 and experiments mentioned only in the text, where the authors report about different sizes of animals. Moreover, as no error bars are shown, it seems that the experiments have not been repeated sufficiently. What I did not understand was why the data in Fig. 5 are not given, as in Fig. 4, as copies/µl? What are the values on the y-axis in Fig. 5? Furthermore, the data shown in Fig. 5 suggest that the effect of 9-cis-RA on Trichoplax was strongest at 1 nM concentration. But why then were the experiments shown in Fig. 4 only carried out at 3 nM and at higher concentrations, where, according to the experiments shown in Fig. 5, rather opposing effects of 9-cis-RA have been observed? 5) The date shown in Fig. 5 are interpreted by the authors as an expression network induced by 9-cis-RA. It remained unclear to me, however, whether the authors imply that 9-cis-RA elicits its effect solely via binding to TaRXR, which in its activated form then activates the expression of its own gene and the genes of the other NRs, or whether they think that 9-cis-RA can also interact with the other NRs at nanomolar concentration. Binding experiments with 9-cis-RA on other purified Trichoplax NRs could be carried out to prove that TaRXR is the only nM-affinity receptor for 9-cis-RA. Another possibility might be that the Trichoplax NRs could form heterodimers. 6) In the section about expression pattern of NRs (reported in the text and in Fig. 5) in Trichoplax, the authors write (line 346) "From the experience we gained by culturing T. adhaerens, as well as from the previous experiments we knew that the culture conditions can dramatically influence phenotype. Having the possible developmental functions of the ancestral NRs in mind, we raised the question whether the expression patterns of the NRs reflect changes in phenotype." To me it is not entirely clear what this statement is supposed to mean. Why wasn't the expression level of different NRs established first on a normal, large enough population of animals with different sizes (i.e. before selecting animals with different sizes)? Is the difference in the expression profile in small and big (large) animals really significant (33% of RXR compared to other NRs in big and 24% in small animals)? And most importantly: what was counted as big and what as small animal? How was the size of the animals determined? Which animal population, the big or small or both sizes, showed an increase to 51% after treatment (line 354)? In the Methods it is written that the animals were treated only during sunny days. Is this an important information as all 9-cis-RA binding experiments have been performed in the dark? I explicitly ask this, because in the following part of the manuscript, it is reported that the feeding conditions have a visible impact on the animals. This statement is not supported by a rigorous analysis, however. Some pictures of animals in different shapes and colours are shown in Fig. 6, which, in my opinion, solely document that the animals can vary, but it remains unclear whether the variation was induced by a certain feeding condition. It also remains unclear whether the poor growth of Trichoplax on Chlorella (Fig. 7), for example, is due to the low nutritive value or composition of that particular algae or simply because this algae is not the preferred food source of Trichoplax or because the algae is better protected against being captured by the animal. In that case, Trichoplax would just eat less. In Fig. 8, I did not understand what exactly was plotted. In some panels it might be the number of animals and in others the area, as there is no legend given. Have the growth experiments been carried out only once for each condition, because there is again no error bar given (Fig. 7 and Fig. 8)? Experimental design: no comment Validity of the findings: no comment Additional comments: no comment

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: TRICHOPLAX ADHAERENS REVEALS A NETWORK OF NUCLEAR RECEPTORS SENSITIVE TO 9-CIS-RETINOIC ACID AT THE BASE OF METAZOAN EVOLUTION Review round: 1 Reviewer: 2

Basic reporting: no comment Experimental design: no comment Validity of the findings: no comment Additional comments: Novotney et al studied an ancestral Trichoplex RXR ortholog and find that it is activated by 9-cis-RA leading to the expression of other nuclear receptors. This is a neat report. To my mind it is the first example of a nuclear receptor ortholog in a basal metazoan being activated by RA. Comments 1. The effect of algal extract on gene expression is very interesting. Phytochemicals are known to activate the estrogen receptor. Thus, I have a bias for phytochemicals also activating RXR. However, phytochemicals could activate pathways that increase the synthesis of retinoids or other novel endogenous molecules that are ligands for RXR. This is a strong alternative possibility. Considering the strong sequence similarity between Trichoplax RXR and human RXR, the extract should affect human RXR action, an experiment that would provide support for Novetney et al. 2. RA binding to RXR regulating other nuclear receptors in Trichoplax is very interesting. Activation occurs at 1 nM with a clear decrease at higher concentrations. Two thoughts: first, assays at 0.5 nM and 0.1 nM would be of much interest. Regulation at these concentrations would increase the impact of this research. Second, a comment on the decreased activity at relatively low concentrations of RA [3nM] is appropriate. It suggests to me that RA is not the endogenous ligand and thus, Trichoplax does not have metabolic pathways to regulate RA homeostasis as is found for steroid homeostasis in vertebrates. 3. I think the Introduction could be better focused on early metazoans, such as Trichoplax, Jellyfish and sponges. The discussion of invertebrates such as C. elegans and Drosophila could be condensed and moved to the final Discussion or eliminated. The strength of this paper is its data on the simplest animals. Trichoplax contains 4 nuclear receptors, in contrast to C. elegans and Drosophila which are more complex and not as ancient. 4. In 2008, I published a paper in BBRC [Trichoplax, the simplest known animal, contains an estrogen-related receptor but no estrogen receptor: Implications for estrogen receptor evolution 375, 623-627 (2008)] analyzing ERR in Trichoplax. As part of my analysis I did BLAST searches that showed that Trichoplax contained orthologs of RXR, HNF4 and COUP. Like Novotney et al, my BLAST searches found that Trichoplax RXR had very strong [unexpected] sequence similarity to human RXR. 5. The strong conservation of Trichoplax RXR ortholog and mammalian RXR deserves some comment. When I was studying ERR in Trichoplax, I was surprised at the strong similarity between Trichoplax RXR and human RXR. The BLAST scores were off the charts. In vertebrates, RXR functions as a heterodimer with several nuclear receptors. The requirement to bind several diverse nuclear receptors would be expected to constrain the RXR sequence. RXR forms heterodimers with Ecdysone receptor. You also cite evidence that jellyfish RXR forms heterodimers with TR. This indicates that heterodimers of RXR evolved early in metazoan evolution, which could explain the constraints on sequence and the strong similarity of Trichoplax RXR and human RXR. At least that is one simple explanation. Moreover, I think there should be a comment about RXR acting as a homodimer instead of a heterodimer, with a transition to heterodimer in the evolution of more complex metazoans. This also implies that functioning of RXR in heterodimers of nuclear receptors was an important event in constraining sequence changes in RXR. Sometime in the future it would be worth determining if Trichoplax RXR forms heterodimers with human TR. 6. Line 95 mentions that an ancestral NR possessing gene regulatory capacity, may have been an unliganded molecular regulator. This is, to me, a reasonable possibility. Do the authors have a reference to this hypothesis? 7. Line 99. Is this a jellyfish? I would like a little more description of the gene and what was found. 8. Line 101 replace homology with identity. 9. Line 38, 125 and numerous other places. I think that orthologs are more accurate than homologs. To me, two genes are orthologs if they are descended from the same gene by speciation. Usually they have the same function, although I think it likely that that some orthologs may acquire a new function over time. As I understand it, the main point of this manuscript is that Trichoplax has a gene that has the same ligand binding activity at human RXR, as well as strong sequence identity. Thus, Novotny et al are proposing that the Trichoplax gene is an ortholog of vertebrate RXR. Also consider replacing homolog with ortholog in other places in the manuscript.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: TRICHOPLAX ADHAERENS REVEALS A NETWORK OF NUCLEAR RECEPTORS SENSITIVE TO 9-CIS-RETINOIC ACID AT THE BASE OF METAZOAN EVOLUTION Review round: 1 Reviewer: 3

Basic reporting: Good. The title should be changed as the data does not show an "endocrine-like network". It shows potential components of a potential network but there is no proof of connectivity and no proof of endocrine-likle function. Experimental design: Good Validity of the findings: Good Additional comments: Review of “Trichoplax adhaerens reveals an endocrine-like network sensitive to 9-cis-retinoic acid at the base of metazoan evolution (#6219)” Thanks to the authors for submitting this paper. The English is excellent and concise, the figures and headings are clear, and the subject interesting. I suggest the paper in accepted for publication with some minor changes. I have attached a copy of the PDF with changes in notes on the document but these are also detailed below. My main criticism of the manuscript is that parts of the discussion (and some of the results section) try to push potential ideas and suggestions of function a little too far. As such some of the conclusions should be softened and some of the rhetoric about potential function should be removed. In addition the title must be changed to reflect the comments made in the conclusions (the data does not show an "endocrine-like network". It shows potential components of a potential network but there is no proof of connectivity and no proof of endocrine-likle function.) See pdf for accurate corrections… L 104 - NR3 should be NR3B (There are probably better refs but this is clear in Vogeler 2014; DOI: 10.1186/1471-2164-15-369) L137 - correction (using becomes implemented in) L139 – please clarify number of animals in each extraction L168 – I think g. Not xg. L209 – GPS coordinates have too many decimal places. L283 – Delete “a high” L294 - These two lists are the same. L308 - With Nuclear receptors I do not find these whole gene identities particularly informative. I would much prefer to see Identity figures for each of the LBD and DBD alongside the whole gene identity, as these give much better idea of functional variability between species. L383 – replace “more affecting” with “influencing” L384 – insert more “… more than” L400 - insert known “arrangement known with” L402 – Dispose? Do you mean display? Or perhaps "...are disposed with..." Comments L423 - I agree but I suppose it depends on the assumed phylogenetic position of the phylum Placozoa (as mentioned above). L438-L458 The evidence you have presented is convincing. However (and you have eluded to this), it is still not proof enough and there is a substantial link missing here in cause and effect relationship. I think you need to soften your stance, clarify exactly what is proven and perhaps suggest what needs to be done experimentally to prove the signalling cascade you are hypothesizing. To clarify... Proven... RXR binds 9-cis-RA. 9-cis-RA exposure increases expression of metabolic gene. Hypothesis (not proven)... 9-cis-RA binds RXR and induces expression of metabolic gene. L460 - This title should be changed in line with comment below (L479) (The data only shows a connection in expression levels upon treatment. It does not demonstrate a regulatory connection or formation of a network) L479-481 - I find this a stretch. The data only shows a connection in expression levels upon treatment. It does not demonstrate a regulatory connection. L507-L510 - The hormone story is perhaps a bit of a stretch (a hormone is a signalling molecule produced by the animal itself; food can mimic a hormone but it isn't a hormone). A more likely hypothesis would be that certain food contains essential nutrients at levels that allow T.a. to grow more efficiently. This is the case for almost all algal feeding inverts and is well characterised in commercial bivalve species. L511-512 This is a really nice (and probably the most important) finding. L512-L516 This is the logical next step but the work in this paper does not go this far. I think you should soften the rhetoric in the second half of this paragraph. Or simply even delete it. L529 - What endocrine-like network? There is no endocrine like network proven. Can you write this instead "In conclusion, the presence of functional NRs in T. adhaerens supports ..."

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: TRICHOPLAX ADHAERENS REVEALS A NETWORK OF NUCLEAR RECEPTORS SENSITIVE TO 9-CIS-RETINOIC ACID AT THE BASE OF METAZOAN EVOLUTION Review round: 2 Reviewer: 1

Basic reporting: - Experimental design: - Validity of the findings: - Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: TRICHOPLAX ADHAERENS REVEALS A NETWORK OF NUCLEAR RECEPTORS SENSITIVE TO 9-CIS-RETINOIC ACID AT THE BASE OF METAZOAN EVOLUTION Review round: 2 Reviewer: 2

Basic reporting: no comment Experimental design: no comment Validity of the findings: no comment Additional comments: This is a much improved manuscript, starting with the new title. They have made numerous edits in response to the other reviewers, and this improves the manuscript. Although the authors decided not to do additional experiments that I suggested, the data and conclusions are sufficient to merit publication; that is, I think this is a useful contribution to the literature. It is a long manuscript covering both molecular studies and feeding studies. This may influence the readership. I hope there will be follow-up studies of experiments suggested in my original review.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: MULTI-SCALE IMMUNOEPIDEMIOLOGICAL MODELING OF WITHIN-HOST AND BETWEEN-HOST HIV DYNAMICS: SYSTEMATIC REVIEW OF MATHEMATICAL MODELS Review round: 1 Reviewer: 1

Basic reporting: Clearly written. Manuscript's structure is appropriate. Experimental design: Goal and justification of this study, as a systematic review, are clearly stated. Methodology of systematic review is well presented. Validity of the findings: Conclusions are well stated. Additional comments: 1. Assuming viral shedding as part of the basic viral dynamics model (line128, page 7; Table 2) is uncommon. A more representative example is found in Perelson 2002, Modelling viral and immune system dynamics, Nature Reviews. 2. Do not use "et al." when there are only two authors; e.g. "Martcheva and Li" instead of "Martcheva et al" (line 134, page 8)—also in lines 140,153, 181, 220, 222, 238. 3. Adding a formulation of model's summary for within host scale models for "HIV evolution" (lines144-149) and "HIV and therapeutic interfering particles" (lines 158-164) would improve the introduction of models (similar to Tables 2, 3, etc.). 4. Note that Saenz & Bonhoeffer (2013) also consider acute, latent and late stages of HIV infection, so this should be mentioned in the corresponding section (lines 177-185). 5. In section "HIV transmission rate as a function of viral load" (lines 212-233), mention the different types of functions that are used (e.g., linear, Hill's). 6. A better reference for the sentence "The viral load (and thus the transmission rate) is high during the acute and late stages of HIV infection while being low during the latent stage" (lines 214-216) is: Hollingsworth et al. 2008, HIV-1 transmission by stage of infection, Journal of infectious diseases. 7. A few typos in References list (e.g., "hiv" in line 376). 8. Models diagrams (Tables 2, 3, 4, 5, 6, 7) should include state variables when stating rates, otherwise they are inconsistent (e.g. "k" is not a rate in the same sense as "d"). 9. The reproduction rate of T related to Tm is missing in the model diagram of Table 5. 10. Note that not all studies considered in the manuscript used the function r*V(t) as implied in Table 8.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: MULTI-SCALE IMMUNOEPIDEMIOLOGICAL MODELING OF WITHIN-HOST AND BETWEEN-HOST HIV DYNAMICS: SYSTEMATIC REVIEW OF MATHEMATICAL MODELS Review round: 1 Reviewer: 2

Basic reporting: 'no comment' Experimental design: 'no comment' Validity of the findings: 'no comment' Additional comments: The manuscript is well organized. Thank authors for writing the manuscript carefully. 1. The manuscript is a review article. The review was done in a systematic way as described in lines 102 and in Figure 2. 2. The authors reviewed 9 articles selected from 89 found in the PubMed database. The inclusion and exclusion criteria were clearly defined in the METHODS section (line #98). 3. The authors summarized the results found in those selected papers and stated in the RESULTS section starting at line # 109. One of the key findings is the within host viral load that enhances between host infection. They also highlighted the impact of super infection, antiretroviral therapy, drug resistance, treatment at early or late stages on HIV infection and prevalence. 4. The authors though saw the increased complexity of the multi-scale modeling, but found significant public health impact of these models. 5. This review should be helpful for further studies to eliminate the long time persisted disease HIV-AIDS. With regards to your concern “regarding the assessment of the experimental design or the validity of the findings were provided”- The authors of the manuscript did not conduct any experiments. So, there was no issue of validation of findings. However, they have followed a systematic review process that I mentioned in my opinion (1,2) above. Based on my findings I have no objection to accept this manuscript for publication.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: MULTI-SCALE IMMUNOEPIDEMIOLOGICAL MODELING OF WITHIN-HOST AND BETWEEN-HOST HIV DYNAMICS: SYSTEMATIC REVIEW OF MATHEMATICAL MODELS Review round: 1 Reviewer: 3

Basic reporting: The authors are making a systematic review of the use of immunoepidemiological mathematical models of HIV. Overall, they found 9 articles where such models were used. Although the paper contains interesting pieces of information, the basic reporting would require modifications. A clear definition of the objectives is likely the most important issue, followed by a need for synthesis and better cohesion between the elements of this paper. Without a clear understanding of the review’s objectives, it was difficult to assess both the relevance and adequacy of the results. I assumed that the authors meant to review how, in the mathematical models, the within-host evolution of HIV was considered to affect the transmission of the disease at the population level. To that regard, it was difficult to find this information in the paper. In terms of cohesion, working on the link between the pieces of information that are reported would help, as the sentences sometimes gave the impression to be “floating”, like bullet points. A clearer definition of the objectives would greatly help in that regard, as the information could be linked back to these objectives throughout the paper and hence give a better sense of direction to the reader. The authors could also consider discarding some of the information that is given, depending on these objectives. In terms of the context, the authors should consider elaborating on the clinical relevance of immunoepidemiological models. Maybe a good way would be to describre models that consider, separately, the within-host or the between-host HIV dynamics, especially the questions they were able to answer but mostly those that cannot be answered because they are not considered altogether. Finding more precise and clinical research questions relative to what is written in lines 77-80 could help improve motivation. Also, is it the first time immunoepidemiological models are reviewed in the context of HIV? As for the Results section, is it possible that some sentences from the Results were statement of facts or data-driven conclusions (e.g. lines 133-134, 159-160, 178-180, 187, 191-192), rather than model elements of the reviewed paper? If this is the case, these sentences would be a better fit in the Introduction. I would avoid dividing the results in very short sections and work on synthesizing the information. Some statements were ambiguous, e.g. sentences like “… can be modeled similar to co-infection (Metzger et al. 2011)… (line 161)” left me to wonder if the referenced authors are the ones stating that it could be modeled, or that they actually included this element in their model and, if so, if they were the only one considering TIPs. I believe important that the authors identify these sentences and make more precise statements. I suggest systematically reporting the frequency of papers having considered which specific model element and why they have done so. Maybe a table with one line per model and check marks for elements considered (columns) could help? Table 1 contained the most interesting information. I suggest elaborating more on the important aspects of it. Consider discarding Figure 1, Tables 2 to 9, which did not bring much to the reading, unless they can be related to an objective. Table 9 could help with motivation (see above comment), but I would also add why it is important to assess these questions. Experimental design: The general methodology seems adequate for a systematic review. However, I was wondering if the authors considered using MeSH terms, from PubMED; this is a great tool to find all terms related to concepts. Also, I suggest looking at other databases, as mathematical models can be underrepresented in PubMED. Validity of the findings: I was unable to relate the discussion to the objectives, for the issues stated above. Additional comments: No further comment.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: MULTI-SCALE IMMUNOEPIDEMIOLOGICAL MODELING OF WITHIN-HOST AND BETWEEN-HOST HIV DYNAMICS: SYSTEMATIC REVIEW OF MATHEMATICAL MODELS Review round: 2 Reviewer: 1

Basic reporting: No comment Experimental design: No comment Validity of the findings: No comment Additional comments: All my previous concerns/comments have been satisfactorily answered. I have no further comments.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: MULTI-SCALE IMMUNOEPIDEMIOLOGICAL MODELING OF WITHIN-HOST AND BETWEEN-HOST HIV DYNAMICS: SYSTEMATIC REVIEW OF MATHEMATICAL MODELS Review round: 2 Reviewer: 2

Basic reporting: The authors made modifications which improved my understanding of the nature of the reported results. Some of my concerns remain. My main concern pertained to the clarity of the stated objectives (see below for details). To this end, I understand that no modification was made by the authors. Consequently, my troubles relating the reported results to the objectives remain. Of relatively lesser importance is the text’s lack of direction. A lot of information is given and it would help a lot if the authors explained, before the Results section, how they chose which information to report and the purpose of reporting this information. I also had a bit of difficulties finding the pieces of information which I thought most relevant to the paper. I leave it to the editor to decide if the latter issues are important enough to be tackled. Objective To be more precise, I struggled understanding the last part of this sentence: “The objective of this study is to conduct a systematic review of multi-scale HIV immunoepidemiological models to infer the synergistic dynamics of HIV prognoses at the individual level and the transmission dynamics at the population level.” Is this what the authors meant: “The objective of this study is to conduct a systematic review of multi-scale HIV immunoepidemiological models to better understand how the evolution of the virus within the hosts could impact the transmission of the disease between individuals, and vice-versa.” ? Please clarify in the text. Results Most of the Results section focuses on technical aspects of the modeling approach. What is the general purpose of presenting this information? Please provide a rationale in the text. In my mind and if I understand the objectives, the authors wished to inform the synergy between within- and between-host dynamics. If this is the case, then model inferences can inform this synergy (Table 1), not the modeling approach. The modeling approach, on the other hand, informs the reader of the validity of model inferences (see suggestion below for details). Although it is quite possible that this was the authors’ intent in presenting the modeling approach, how the results are presented does not help the reader make the appropriate assessment of inferences. Accordingly, here are some modification suggestions: - Most models have a great number of underlying assumptions (e.g. homogeneous populations, deterministic/stochastic nature of a relationship, etc.). Please mention why the specific model assumptions were chosen to be reported in the objectives. As an example, you could write: “Inferences from mathematical models are sensitive to model assumptions. To enable a better assessment of the validity of inferences, we will report the model assumptions that we believe had a larger impact on model predictions.” - In the Results section, it would be advisable not to separate data-driven facts, model assumptions and model inferences, and make it clear which is which. As an example based on lines 275-279 (this is not meant to be accurate!): “There is evidence suggesting that deploying therapeutic interfering particles in even a small proportion of infected individuals reduces the prevalence of HIV to low levels. (ref=longitudinal study) This is thought to be due to TIPs’ ability to transmit between hosts and and target high-risk groups. (ref=source of this opinion) In one immunoepidemiological model, this was modelled by assuming that the susceptible population infected by TIP carriers become TIP carriers themselves, and those have one-half the transmission rate of a non-TIP carrier.(ref=model paper) This assumption was supported by clinical data. Results from analyzing this model suggest that TIP may reduce the challenges of ART therapy and vaccines, and can be complementary to both.(ref=model paper) This is due to [particular aspect of the model dynamics] and adds to the current knowledge ...” - If I understood your objectives correctly, I suggest describing the most important information relative to your objectives in the text rather than in a table (e.g. Table 1). I would put less relevant technical aspects of the modeling approach in Appendix. Experimental design: no comment Validity of the findings: no comment Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: AN EXPLORATION OF STRATEGIES USED BY DRESSAGE HORSES TO CONTROL MOMENTS AROUND THE CENTER OF MASS WHEN PERFORMING PASSAGE Review round: 1 Reviewer: 1

Basic reporting: No comment in general, please add abbreviations for table in legend Experimental design: No comment Validity of the findings: No comment Additional comments: This paper studies the biomechanical variables which influence pitching moments around the COM in passage in elite dressage horses. It uses comprehensive recording and analysis to understand the background for the high biomechanical skills of three elite dressage horses. There are two points regarding the study object and presentation of the results which should be considered: 1. “balance control”: In the abstract, introduction and discussion the term balance or balance control, balance management is used. You also defined this term in the abstract as the ability to control nose up/nose down pitching moments. This might be a quite limited part of the general locomotor balance of the horse. To my knowledge the FEI does not exactly define this term “balance”. While this study focusses on the sagittal-plane motion, the transversal plane movements might be of even more importance especially during slow gaits. As balance might mean generally “a state of equilibrium“ and the capacity to maintain postural stability, it might be nice to discuss these aspects in your discussion in more detail. 2. Results: Table 1: Please provide the explanations for the abbreviations in the table legend.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: AN EXPLORATION OF STRATEGIES USED BY DRESSAGE HORSES TO CONTROL MOMENTS AROUND THE CENTER OF MASS WHEN PERFORMING PASSAGE Review round: 1 Reviewer: 2

Basic reporting: Please see the annotated PDF for minor comments on phrasing. I find there are limited references within the article, particularly in the introduction. It is not mentioned that this topic is understudied, so I am surprised that there are not more citations. I don’t believe in including irrelevant citations for the sake of it, but some mention of limited references, or inclusion of more support for discussions of trot and pitching moments in other movements may be beneficial. This would also help to demonstrate how this research fits with the broader field. It would be more logical to define pitching moments in the introduction, as it is an important aspect of the study, as opposed to the discussion (Lines 219-221). The figures are clear and appropriate to the article; however, I’m not sure why they are subplots of the same figure. It would be clearer to have two distinct figures, and these should also be referred to specifically in the text. Figure 1 is referred to, but not 1A or 1B. Experimental design: There are minor phrasing comments within the annotated PDF to clarify the Materials and Methods section, but overall it contains sufficient detail. The research fits within the Aims and Scope of the journal as stated on the website, and the research question is fairly well defined. The statement “some dressage horses fail to reach the highest levels…” (Lines 65-67) is interesting and a nice rationale for the study, but is buried within the introduction. The authors use the term “elite” as well as “highly-trained” dressage horses. I think there needs to be a clear definition of these terms. The use of only three horses seems very small for “highly-trained” dressage horses, but it more acceptable if “elite” is referring only to Olympic competitors. There should also be acknowledgement of the small sample size before the end of the discussion, which could be accomplished by defining the criteria for inclusion. Validity of the findings: I believe the methods and analysis are sound, and the results presented are reliable. I particularly like the supplementary video. The discussion and conclusions are appropriate to the results of this study, but there are aspects of the discussion that I think could be improved. The first sentence of the discussion states that dressage “…involves maintaining the trunk in an uphill (nose-up) orientation” (Line 201). This is a key factor of the pitching moments discussed in the paper, but is not mentioned before the discussion. This should be brought to the attention of the reader in the introduction, and would also assist with identifying the purpose of the study. The manuscript overall is focused on healthy horses performing a difficult movement; the mention of disease (Line 211) seems off topic and unnecessary. I think it should be removed. I don’t think it’s necessary to mention that there are no significant differences between breeds when all the individuals were the same breed (lines 262-265). The final sentence of the conclusions refers back to lines 65-67, which I have already stated is where I believe the key rationale for this study to be. But, as it’s not mentioned within the discussion, this last sentence seems oddly placed. I recommend highlighting this more within the introduction and discussion. The discussion overall could be brought together more effectively. It reads a little disjointed, each result is discussed separately, rather than passage overall. There could be greater comparison between passage and trot, as this is briefly mentioned but not clearly compared. For example, lines 227-229, the comparison here is not clear (I’m not sure what “controlled by manipulating both […] GRF moments” means in relation to the trot). There is also little mention of the three balance strategies that are highlighted earlier in the manuscript. They are included in Table 1, but then not discussed. This would be a good way to bring together all the information from the discussion to an overall mechanism for balance in the passage. Additional comments: Overall, I think this is a well written, PeerJ appropriate manuscript. The methods are ethical and appropriate, the data is sound, and raw data is provided as supplementary information. There are areas that could be improved to demonstrate the research question that this manuscript is aiming to address.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: SLOWER EEG ALPHA GENERATION, SYNCHRONIZATION AND “FLOW”—POSSIBLE BIOMARKERS OF COGNITIVE IMPAIRMENT AND NEUROPATHOLOGY OF MINOR STROKE Review round: 1 Reviewer: 1

Basic reporting: It is not immediately obvious in the abstract that stroke patients where tested at both sub-acute and chronic stages. Please add additional detail to make this clear. The authors have used "ɑAVG" in equations for calculation of both weighted average alpha frequency (line 173) and probability density estimate (line 182). Is this an error? Groups are represented by different shades of blue in all figures, and to my eyes, the light blue used for the control group is almost too faint to comfortably see. Consider using a different color scheme. Experimental design: No comment Validity of the findings: Age comparisons between groups were made for stroke patients in the sub-acute, but not chronic stage. Was the age of stroke patients at follow-up in the chronic stage still matched to controls? Potentially important given well described effects of advanced age on EEG alpha (as mentioned in the Discussion, lines 452–458). Most significantly for the the coherence measures, it seems likely that volume conduction is a major biasing factor. Also a potential issue is the use of a common reference, in this case, the average of C3 and C4 (for example, see Nunez et al. 1997 Electroencephalogr Clin Neurophysiol 103:499–515; Nolte et al. 2004 Clin Neurophysiol 115:2292–2307). It is recommended the authors consider using analytical methods that limit these confounds (e.g. imaginary coherence). Additional comments: No further comments.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: SLOWER EEG ALPHA GENERATION, SYNCHRONIZATION AND “FLOW”—POSSIBLE BIOMARKERS OF COGNITIVE IMPAIRMENT AND NEUROPATHOLOGY OF MINOR STROKE Review round: 1 Reviewer: 2

Basic reporting: LANGUAGE: Mostly good usage of English language throughout. Some errors e.g. “neurophysiological” wrongly used in place of “neuropsychological” (lines 88, 119); “above the overall” (twice, in line 368) should be changed to “overlying the”, or similar; “for the” is missing from “compensate for the lesion” (line 57 & elsewhere). The first sentence of Results in the Abstract (lines 43-46) is too long and complex, and needs to be separated into two sentences (at least). INTRO, BACKGROUND, CONTEXT & LITERATURE: The Introduction provides a reasonable summary of literature relevant to this study, although with some exceptions and some shortcomings. In particular, the chief focus of this study is on EEG alpha activity. However the authors state in the Intro that “alpha and theta frequency could be of interest for the screening of post-stroke cognitive impairment” (line 116). Why then is theta activity not analyzed in more detail in this study? In addition the Schleiger et al (2014) study is cited, but the authors overlook the finding from that study that delta/alpha ratio (DAR) was found to correlate with post-stroke cognitive outcomes. Why was DAR not analyzed in this study? The authors also overlook that the Schleiger et al (2014) found alpha power from all electrodes (not frontal or posterior specifically) to correlate with post-stroke cognitive outcomes – but it seems that this study only analyzed frontal or posterior electrodes. The Intro and manuscript also is lacking a citation of Schleiger et al (2017) which reports findings that are more recent and relevant to this manuscript: including re peak alpha frequency and slowing of alpha after stroke. So more justification is needed in the Intro for the authors’ choice of EEG parameters. Or – more parameters (e.g. DAR, or lower-frequency / slower alpha range) could be analysed. Line 83: post-stroke cognitive impairments are not always, and not only, consequences of frontal “impairments”/damage specifically. For example, damage to other brain regions – e.g. medial temporal, parietal cortex – can also results in cognitive impairments. STRUCTURE: appears appropriate. FIGURES: appear appropriate; possibly too many (7) figures? RAW DATA: I’m not aware as to whether or not the raw EEG data are available? Experimental design: SCOPE: appears to be within journal scope. RESEARCH QUESTION: I think the question and primary aim is not very well defined or identified. We have stroke and control samples. From the stroke sample we have two assessment times – mean 9.7 days and mean 13.7 months after stroke. Was EEG recorded at both times? This is ambiguous (line 151). Stroke symptoms (by NIHSScale) and cognitive function (by MoCAssessment) were assessed at both times. It remains unclear as to whether the focus is on: controls v stroke? Longitudinal changes in stroke EEG? Correlations between (early) stroke EEG and outcomes (MoCA / NIHSS?) from stroke? Instead there seems an over-emphasis on EEG analyses and parameters. I believe the authors need to more clearly identify and state what is their primary research question and aim; also what is the secondary (and perhaps tertiary) question of aim. STANDARDS OF INVESTIGATION: No issues (apart from lack of focus of primary question and aim). METHODS DETAIL: Generally adequate, although a few missing details. Was EEG recorded at both assessment times? - this is ambiguous (line 151). Was (C3+C4)/2 the online, or offline, reference technique (or both)? In any case – this is unusual. Common average reference is mentioned in a later section (line 202), but was this applied for ALL EEG analyses and parameters? Was it average of all 19 electrodes used, or average of only the 8 electrodes analyzed? Were ONLY the 4 frontal and 4 posterior electrodes analysed (line 164)? – the authors say “particularly” but I think they mean “only”? Was relative power calculated relative to 1.1 – 50 Hz power, or other range (line 168-9)? For each frequency band was power summed or averaged? What was the frequency resolution of the FFT output? Validity of the findings: Results seem inconclusive, mainly because they are from a small, specific sample of stroke patients: 9 patients with Right middle cerebral artery (RMCA) stroke, who had minor strokes (NIHSS < 5). This is acceptable for this journal, but this disclaimer needs to be stated. The “benefit”, or implications, for the literature could be proposed. For example – these findings from a (small) sample of RMCA patients, could not be extrapolated to LEFT MCA cases (even if the RMCA sample size was large). DATA: the main issue is the small and specific nature of the sample (see above). CONCLUSIONS: Need to be tempered (i.e. disclaimer added) as appropriate given that small, specific sample of stroke patients - 9 patients with Right middle cerebral artery (RMCA) stroke, who had minor strokes (NIHSS < 5) – were studied. Some conclusions or interpretations seems to general or perhaps speculative – e.g. re “alpha flow” (lines 375-7; 471; 59) and lines 449-451. Some speculation is ok (see below) but this should be presented as speculative proposal(s) rather than “firm, conclusive or final”. Discussion paragraph (lines 452-8) seems quite irrelevant and could be deleted (or shortened). SPECULATION: The conclusions and claims of the authors, as they stand, need disclaimers or caveats (as recommended above) and thus should be considered relatively speculative at this early stage. Additional comments: I encourage the authors to revise the manuscript as recommended - this will result in an improved version of what is a quite interesting study and manuscript.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: SLOWER EEG ALPHA GENERATION, SYNCHRONIZATION AND “FLOW”—POSSIBLE BIOMARKERS OF COGNITIVE IMPAIRMENT AND NEUROPATHOLOGY OF MINOR STROKE Review round: 2 Reviewer: 1

Basic reporting: No comments Experimental design: No comments Validity of the findings: With regard to volume conduction and use of a common reference biasing coherence measures: I am willing to accept that the findings of this study are still valid without altering the analyses. Nevertheless, the potential influence of these confounds is still a limitation that needs to at least be acknowledged in the Discussion. Additional comments: Nil

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: SLOWER EEG ALPHA GENERATION, SYNCHRONIZATION AND “FLOW”—POSSIBLE BIOMARKERS OF COGNITIVE IMPAIRMENT AND NEUROPATHOLOGY OF MINOR STROKE Review round: 2 Reviewer: 2

Basic reporting: Reporting is generally satisfactory. I recommend just two improvements: [1] The second-last sentence of the Discussion (lines 499-504, tracked version) is simply too long, and thus it also becomes potentially confusing. Please break this down into at least two sentences. [2] The revised Aim at end of Intro (lines 141-145) is improved, however it remains too long and thus, the aims seem to diffuse or lacking adequate focused. Please deconstruct this into at least two sentences - the first should contain the principal/ primary aim, and the second sentence can state secondary/ supplementary aims. Experimental design: No further comments. Validity of the findings: No further comments. Additional comments: I feel the authors have done a reasonable job in revising their manuscript in response to comments from myself and the other reviewer. If the two very long sentences (at end of Intro and end of Discussion) can be revised and improved (following my comments above) i believe this manuscript will then be suitable for publication. I wish the authors well in their future research in this field.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: PREDICTING STIMULATION-DEPENDENT ENHANCER-PROMOTER INTERACTIONS FROM CHIP-SEQ TIME COURSE DATA Review round: 1 Reviewer: 1

Basic reporting: I think that in general the article is well written, in clear professional English. Literature is in general cited sufficiently (with one exception described below). However, some details are missing and especially the figures need improvement. Major points: My main point of criticism here is about the quality of the figures and the lack of explanation about what they are showing. The reader basically has to make too many assumptions to be able to interpret them with confidence. - Fig. 1: what are the ranges shown? Center of the cluster +/- SD? What are the numbers above each plot? The number of regions in the cluster? What is the Y axis? Log fold change? Why are 12 clusters shown in each subplot? Another problem is that many clusters seem practically indistinguishable from each other (see for example the four top plots in Fig. 1a). Should these really be considered as different clusters? They might be merged, I feel. - For Fig. 2e, people will wonder why the density is 0 for the smaller distances. I guess this is because of the way intergenic enhancers were defined here. It would be better to indicate this in the figure; for example: mark the region (+/-300bps?) where enhancers would be filtered out. - For Fig. 3: the colors in Fig. 3a are not clear. Adding gene symbols to the genes might help the understanding. Fig. 3b-g: Adding some distance between the training and test panels would make things more intuitive. An explanation is needed about what “PolII”, “ER”, etc on top of each plot means, as well as what “NB” stands for, and what is the meaning of “distance” and “data”. Please also explain what the error bar means. At present, all of this has to be guessed by the reader. There appears to be a “3” in Fig. 3f, third panel, which should be removed. Finally, at present 5 panels are shown for training data, and only 1 for test data. It would make more sense to show performance on the test data in more detail, rather than on the training data, as this performance gives a better idea of the general performance on new data. Minor points: - In November 2016, several papers (see Cell Volume 167 Issue 5, including Spurrell et al., Cell, 2016; Javierre et al., Cell, 2016; etc) were published by the BLUEPRINT Consortium/ International Human Epigenome Consortium (IHEC) describing analysis of promoter-enhancer interactions using promoter capture Hi-C in multiple cell types and conditions. I think a reference to these studies and their expected impact in the future should be added to the Introduction. - I think that no reference was given for MACS. This should be added. - p4 line 157. “The method is similar to k-means but can achieve much better optimisation of the k-means objective function than the standard EM algorithm.” As far as I understand k-means and the EM algorithm are two different things, and k-means does not use EM. I would recommend removing this sentence, or providing a citation that supports this statement. - P8 line 320: what are “0.66, 0.47, and 0.35”? Is this correlation? A more clear explanation is necessary. Experimental design: I think that the paper described original research, with a well defined and clear problem and aim. The computational methodology is described in sufficient detail and is quite interesting. However, there are some issues about the newly produced ChIP-seq data (see below). Major point: - New ChIP-seq data (ER-alpha, H3K4me3, and H2AZ) was produced, but details about the experimental procedures are completely missing from the Materials and Methods section. How were antibodies obtained? How about sequencing libraries, buffer solutions, concentration, cell counts, etc? Raw data appears to have been submitted to NCBI GEO, however, the data is still set to private, and no reviewer URL was provided, so I can not comment on the data itself. Minor points: - Were control samples produced for the ChIP-seq data? If no, some discussion about how this could influence the downstream analysis would be useful. Differences in chromatin accessibility can result in non-specific peaks which are hard to remove without using control samples. Some ChIP-seq peaks might therefore simply reflect open chromatin structure, and promoter-enhancer correlations could reflect similar changes in chromatin accessibility, rather than purely binding by ER-alpha or Pol2-II. - In the Materials and Methods section (p5 line 201~) the assumption is made that each enhancers regulates a single gene. I think that this is in practice a reasonable assumption, as it becomes computationally impossible to consider more combinations of enhancer-to-promoter interactions. However, it would seem quite reasonable and doable to add the case where an enhancer does not regulate any of the genes to the model. It would be quite interesting to see what happens if this case (“I(0)_j”?) is included. Does it change the assignments? Will you end up with many enhancers having no apparent target promoter? Validity of the findings: In my opinion the findings are valid, though they are not so easy to interpret with confidence from the figures (see comments above). Improving the quality of the figures are their explanation will improve this point. Additional comments: Some additional (brief) discussion of the approach and results might be of interest for people working on related topics. - Currently you are considering only the correlation between consistent features (for example: ER-alpha at the promoter vs ER-alpha at the enhancer). However, it might be possible that a feature at the promoter is better correlated with a DIFFERENT feature at the enhancer (example: Pol-II at the promoter vs ER-alpha at the enhancer). - Time-lagged correlation has been used for biological network inference in the past. Here, you are not considering this. Some discussion on this would be nice (e.g. would this be computationally difficult? Would this be biologically relevant here?) - (Somewhat related to the above) H3K4me3 levels are typically low at enhancers. Leaving out this data seems to not affect accuracy very much. Other modifications (H3K27ac?) might have been more promising. Some discussion could increase the interest of the publication.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: PREDICTING STIMULATION-DEPENDENT ENHANCER-PROMOTER INTERACTIONS FROM CHIP-SEQ TIME COURSE DATA Review round: 1 Reviewer: 2

Basic reporting: The authors have developed a Bayesian classifier model which integrates genomic distance with a correlation of time-course ChIP-seq data to predict physical enhancers-promoter interactions. The article is well written, and includes sufficient references in introduction to demonstrate the research goal. The methods are clearly described. However, the legend of each figure should be written more carefully. For example, the definition of the shaded regions indicate in Fig 1, the error bars in Fig 3, and the number of data points in each panel in Fig 3, should be explained in legend. Please refer MACS in the references (line 127). Experimental design: The proposed model is evaluated using various data including ChIA-PET, Hi-C and GRO-seq data, while the scope of investigation performed is limited in enhancers-promoter interactions of stimulation-dependent genes for estrogen in MCF7 cells. (Minor comment) Because MACS version 1.4.2 is outdated, the authors should use MACS2 for peak-calling, while the conclusion of the article may hold. Validity of the findings: The conclusion of the article is robust in the data used for evaluation. However, the article focuses the interactions between promoters and intergenic enhancers only, while there are quite a few “genic” enhancers in the genome. Enhancer-enhancer interactions also exist. The discussion is necessary about how the proposed model works for these kinds of interaction. Furthermore, because this method can capture stimulation-dependent genes only, it should be estimated what percentage of interactions were captured by this model against all interactions captured by ChIA-PET assay. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: PREDICTING STIMULATION-DEPENDENT ENHANCER-PROMOTER INTERACTIONS FROM CHIP-SEQ TIME COURSE DATA Review round: 1 Reviewer: 3

Basic reporting: Complete lists of predictions should be provided as Suppl Tables. Some texts in the figures are very small to read. Experimental design: no comment Validity of the findings: no comment Additional comments: Dzida et al. developed a method to predict enhancer-promoter interactions from ChIP-seq time course data. They used ChIP-seq data of ERalpha, Pol II, H2AZ, and H3K4me3 at eight different time points after E2 stimulation in MCF7 cells, and showed, by comparing with GRO-seq for validation, that the method is able to successfully predict distally regulated target genes. In short, I find the use of time course data of several ChIP-seq experiments significantly improve the prediction of enhancer-promoter interactions in genes regulated by ERalpha. However, there are some issues the authors should address in this manuscript. 1. Pg 8, Fig 1. There is no description about blue and red lines in the figure legend. And also, why are there 12 graphs in each panel? What are the numbers at the top of each graph (I guess these are number of instances, but they are not explained in the figure legend.)? Please elaborate more on these points. 2. Pg 8, Line 313. The authors says “..., there is a substantial overlap in the positive and background distance densities due to a large separation…” (in Fig 2e). I don’t see the substantial overlap in this graph. Please clarify or explain in more detail about this. And also, what is the unit for distance (x-axis)? The numbers, especially for the exponential part, are too small to read. 3. Suppl Figs should not be numbered as 1, 2, 3, ..., but as s1, s2, s3, .... 4. Suppl Figs s1, s2, s9, s10, and S11 are not mentioned in the main text. 5. Pg 5, Line 252. I guess ‘TPR’ means ‘True Positive Rate’. Write it out in full the first time. 6. Pg 10, Fig 3a. It is hard to discriminate green and gray lines. What is the difference in thickness of these lines? X-axis must be genomic coordinate, but the unit is not clear in the figure. 7. Pg 10, Fig 3f. What is ‘3’ at the lower right corner of the ‘Pol II, ER’ panel? 8. Suppl Figs s3-6. Why are the chromosomes divided into odd and even ones? 9. Please provide complete lists of predictions as Supplementary Tables (for example, those mentioned in ‘Validation of ER-regulated target gene predictions’ in Pg 11).

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: PREDICTING STIMULATION-DEPENDENT ENHANCER-PROMOTER INTERACTIONS FROM CHIP-SEQ TIME COURSE DATA Review round: 2 Reviewer: 1

Basic reporting: The issues raised by me (and also by the other reviewers) been addressed by the authors. Especially, the clarity of the figures has been strongly improved. Experimental design: The authors have added a detailed description of the experimental procedures of he ChIP-seq data. I think that the manuscript is acceptable for publication now. Validity of the findings: Thanks to the improved quality of the figures and explanation of the methods, the validity of the findings can be more objectively judged. The findings of the paper are valid and suitable for publication. Additional comments: I appreciate the efforts made in improving the quality of the manuscripts. The paper us suitable for publication now.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: PREDICTING STIMULATION-DEPENDENT ENHANCER-PROMOTER INTERACTIONS FROM CHIP-SEQ TIME COURSE DATA Review round: 2 Reviewer: 2

Basic reporting: In the revised manuscript, the authors have added new experiments and explanations, and some of my concerns were satisfied, but the authors should add the discussion about the points below for publication of this manuscript. I agree with the authors that "complementary approaches to infer enhancer-promoter interactions by exploiting readily available sources of genomic data, such as ChIP-Seq and RNA-Seq data, are of interest." However, the advantages and the limitations of the proposed method against chromosome conformation capture techniques are not clearly discussed. Experimental design: No comment Validity of the findings: ・The line 445-447 “Our model can therefore serve as a complementary approach to chromosome conformation capture techniques and offers insight into context-specific, and cell-type specific transcriptional regulation.” This method can only capture the interactions between the promoter and the intergenic enhancer for stimulation-dependent genes, while ChIA-PET and Capture Hi-C can obtain all types of interactions. Moreover, this method used ten time points for each TF, resulting in the requirement of many ChIP-seq samples, which is not cost-effective and is also technically challenging. Such limitation should be discussed in Discussion section. ・The discussion about the minimum time points required the proposed model is also necessary. ・The authors’ comment “We also limited attention to enhancers outside of gene regions. This was mainly due to the nature of our data, since we are using Pol-II as one of our features and this mark is found across the whole body of transcribing genes.” Is this a limitation of the proposed method, or the data used (Pol2)? If there is a dataset of other TF, the genic enhancers can be investigated? ・The definition of “context specific enhancer-promoter interactions” in the title and the abstract is ambiguous, and no explanation in the manuscript. If this means “stimulation-dependent enhancer-promoter interactions”, the authors should replace it. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: PREDICTING STIMULATION-DEPENDENT ENHANCER-PROMOTER INTERACTIONS FROM CHIP-SEQ TIME COURSE DATA Review round: 2 Reviewer: 3

Basic reporting: no comment Experimental design: no comment Validity of the findings: no comment Additional comments: In the revised manuscript, the authors properly responded to all the points raised in the initial round of the review. I don't have any further concerns, and would like to recommend its publication in PeerJ.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ADVANCES IN THE STUDY OF NODAVIRUS Review round: 1 Reviewer: 1

Basic reporting: In overall, the manuscript was written with clear and unambiguous English. There are some points that need to be addressed. The references were sufficient. Experimental design: This manuscript is a review article. Validity of the findings: This manuscript is a review article. Additional comments: Advances in the study of nodaviruses This review manuscript described nodaviruses in various aspects including transmission, classification, immune response and diagnosis using molecular and immunological techniques. Overall, the manuscript thoroughly explain in detail; however, there are some points that should be added for completeness. The whole M.S. should be edited by native English speaker. 1. Line 86: The RGNNV has been shown to infect tilapia (Keawcharoen,J, et al., J. Fish Dis. 38, 49-54 (2015). This information should be added. 2. Line 293: The monoclonal antibody specific to MrNV was reported (Wangman et al., Dis. Aquat. Org. 2012, 98(2):121-31). The information based on this report should be mentioned. 3. Line 143: What is a well-established cell line? The detail is needed. Minor points - The “Mr” in “MrNV” should be italicized throughout the manuscript. - Lines 27, 207: use “Histopathology..” - Line 75-76: use “…….infected fishes. Neither alpha-nor beta-nodavirus………(NaveenKumar et al., 2013). Another type of nodavirus includes……. - Line 78: use “…PvNNV should be categorized…” - Line 79: use “…compare with that of both….” - Line 165: use comprise of - Line 167: use “Both of which ….” - Line 213: use “lymphocytes” - Line 225: use “…with that of members…” - Line 246: italicize “in situ” and use “..nested RT-PCR…” - Line 274: use “..higher than that of …” - Line 279: use “…lower than that of …” - Line 422: change “vertebrate” to “fishes”

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ADVANCES IN THE STUDY OF NODAVIRUS Review round: 1 Reviewer: 2

Basic reporting: This review is well-written. Literature should be implemented. Experimental design: Not applicable. Validity of the findings: Not applicable. Additional comments: The manuscript is a review about Nodavirus. Paper is well written and clear. Respect to the first statements about recent reviews in the field there are some good reviews not included covering distinct aspects: Doan QK, Vandeputte M, Chatain B, Morin T, Allal F. Viral encephalopathy and retinopathy in aquaculture: a review. J Fish Dis. 2017 May;40(5):717-742. Costa JZ, Thompson KD. Understanding the interaction between Betanodavirus and its host for the development of prophylactic measures for viral encephalopathy and retinopathy. Fish Shellfish Immunol. 2016 Jun;53:35-49. Reshi ML, Su YC, Hong JR. RNA Viruses: ROS-Mediated Cell Death. Int J Cell Biol. 2014;2014:467452. Hong JR. Betanodavirus: Mitochondrial disruption and necrotic cell death. World J Virol. 2013 Feb 12;2(1):1-5. Chen YM, Wang TY, Chen TY. Immunity to betanodavirus infections of marine fish. Dev Comp Immunol. 2014 Apr;43(2):174-83. Chaivisuthangkura P, Longyant S, Sithigorngul P. Immunological-based assays for specific detection of shrimp viruses. World J Virol. 2014 Feb 12;3(1):1-10. But also other reviews related to viral diseases, immunity, etc. in aquaculture species during the last decade. Thus, the starting point of this review is not adequate. The topic is certainly interesting and merits further research since nodaviruses are actually producing probably the largest economic loss in the productive sector. Regarding insect nodaviruses, there is only scarce explanation and this could be either deleted from the paper or reinforced with further information (detection, immunity, vaccines…). Comments: - Line 100. The number of infected fish species is much larger. Please, revise it. - Line 104. Starting from this point include the latin name of all animal species in the text. - Line 143. The cell line name is lacking. - Line 320. This reference in not appropriate for this. - Line 340. Classification of Mx as an AMP is not adequate and consistent. This belongs to the type I IFN response rather to the AMP repertoire. - Section related to the immunity, both fish and prawns, needs to be improved and better organized. It is not clear whether the cited studies area about gene expression or functional and this is an important issue to be denoted. References in this section need to be further revised and better selected. For example, important references should be included: doi: 10.1016/j.fsi.2009.11.008. doi: 10.1016/j.dci.2012.02.007 doi: 10.1016/j.fsi.2011.03.018

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ADVANCES IN THE STUDY OF NODAVIRUS Review round: 2 Reviewer: 1

Basic reporting: The English is clear and unambiguous. Experimental design: The review included many updated data. Validity of the findings: The conclusion was drawn as shown in Fig. 1. Additional comments: All of the questions were properly responded. I have no further comments.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ADVANCES IN THE STUDY OF NODAVIRUS Review round: 2 Reviewer: 2

Basic reporting: The paper is OK. Experimental design: NA Validity of the findings: Ok Additional comments: The manuscript has been improved accordiingly.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROBE-BASED QRT-PCR METHOD TO PROFILE IMMUNOLOGICAL GENE EXPRESSION IN BLOOD OF CAPTIVE BELUGA WHALES (DELPHINAPTERUS LEUCAS) Review round: 1 Reviewer: 1

Basic reporting: No comments Experimental design: No comments Validity of the findings: 1. Although the present system shows some potentials to evaluate health conditions of captive beluga whales, it is better to combine other clinical pathological analysis results (e.g., complete blood count (CBC), serum chemistry panel) with the qPCR results to validate the system. 2. The limitations of the present study should be discussed. Changes on RNA level don't guaranty protein changes, especially when the gene changes a little. 3. The rationale selecting the 6 specific genes remain unclear. Both interleukin and respective receptor are crucial for downstream signaling pathway activation. Why some interleukin genes were selected, while a specific interleukin receptor was examined? Additional comments: In the present study, the author investigated several inflammatory genes in captive beluga whales by quantitative RT-PCR. Generally, this is an interesting study and important to whale health monitoring in aquariums. Several other concerns need to be addressed before publication. 1. Animal backgrounds including place, age, sex and health conditions need to be included in a separated table, which is more straightforward to the readers. 2. The term of "A novel universal probe library quantitative reverse transcription polymerase chain reaction" in the title was confusing which needs to be revised. It should be careful to define the library as a universial library given its application area remain to be determined.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROBE-BASED QRT-PCR METHOD TO PROFILE IMMUNOLOGICAL GENE EXPRESSION IN BLOOD OF CAPTIVE BELUGA WHALES (DELPHINAPTERUS LEUCAS) Review round: 1 Reviewer: 2

Basic reporting: 1. Line 65, ‘the short half-lives of leukocyte gene transcripts encoding surface and secreted messengers’ can be replaced with ‘the short half-lives of leukocyte biomarkers’. 2. Line 70, ‘mRNA level’ should be plural. 3. Include the full name of cDNA at Line 93 instead of Line 118. 4. Line 124, Cq is the abbreviation of crossing point. 5. Line 134, add ‘the’ before ‘target gene’. 6. Line 152, “NV of IL-12 in sample A-I was 3.03, demonstrating that the transcript level of IL-12 was elevated.” The authors need to say, “NV of IL-12 in sample A-I was 3.03 and that of the median was XXX, demonstrating that the transcript level of IL-12 was elevated.” Make similar changes for other genes. 7. In Figure 1, does the error bars represent standard deviation? Experimental design: 1. Immunological gene expression changes mainly in leukocytes after exposure, why the authors used whole blood cells instead of leukocytes? The authors need to validate the use of the whole blood cells or to use isolate the leukocytes. 2. There are commercially available ELISA kits to detect these cytokines, and ELISA doesn’t require blood cell separation, cell lysis or qRT-PCR. It is possible that those kits for human or mice cannot detect the cytokines from beluga. There could be one reason why people want use qRT-PCR instead of ELISA. Have the authors tested if the commercial kits work for beluga? 3. While it is clear that the qRT-PCR method is able to measure the expression levels of the genes, it fails to demonstrate its capability to evaluate the health status of the whales. 4. Line 138, ‘Elevated expression levels of IFNγ and decreased levels of TNFα in C-II were suggested as normal responses to social stressors in the environment, as previously observed in a killer whale study.’ How about other whales? How do the authors know they don’t suffer from social stress? Also, the authors took 4 samples from the same whale, why it is seen in only one sample. Validity of the findings: 1. Did the authors observe normalized value outliners in healthy samples, too? Please explain if there is. 2. The authors observed seven outliners but can only explain four of them, please explain why they thought it was enough. Also, if the elevated IL-12 expression in A-I whale suggested activated immune response, why they failed to observe its elevated expression in other 3 unhealthy samples. Same questions for IFN-γ, IL-2Rα, and TNF-α. 3. The authors used normalized value (NV) in this work. However, this term is not commonly used in qRT-PCR analysis. I suggest using ‘relative expression’. Why the authors define NV as Log2(ETCqT/Geomean (ER1CqR1 & ER2CqR2))? 4. What does ‘threshold value’ in Table 1 mean? How did the authors calculate it? People use the threshold cycle (Ct) in qRT-PCR analysis, but the ‘threshold’ here is clearly not Ct. 5. Line 70, it is real-time PCR but it doesn’t enable real-time gene expression measurement. The assay is off-line. 6. Line 144, why the authors use the median of NVs instead of the mean? Additional comments: The paper described a method to quantify immunological gene expression in whole blood cells of beluga whales. The authors extracted mRNAs from the blood cells of healthy and unhealthy whales and compared the expression of six relevant genes. The authors attempted to find the correlations between the change of certain gene expression and health conditions. The paper is well organized, and easy to read. Probe-based qRT-PCR, such as Taqman, is a widely used method to evaluate gene expression, so the authors cannot claim it is a novel method. The authors need to avoid the use of 'novel' unless they can prove its novelty. Additionally, the authors need to improve the qRT-PCR analysis part and the discussion is weak. 1. Line 102 and Line 112, add the address information of the manufacturers. 2. Line 112, the authors need to put the data in supporting information. 3. Line 144, ‘Gene transcript levels were used to establish three arbitrary categories: IL-2Rα (highly transcribed; median NV < 5), IL-10 and TNFα (moderately transcribed), and IFNγ, IL-4 and IL-12 (lowly transcribed; (median NV > 8).’ The sentence can be removed. 4. Line 152, “NV of IL-12 in sample A-I was 3.03, demonstrating that the transcript level of IL-12 was elevated.” The authors need to say, “NV of IL-12 in sample A-I was 3.03 and that of the median was XXX, demonstrating that the transcript level of IL-12 was elevated.” Make similar changes for other genes. 5. SYBR Green was once widely used in qRT-PCR, but nowadays more and more people use probe-based Taqman. The Roche’s universal probelibrary is a new method, but it was not first introduced in this work. 6. Line 215, the most commonly used qRT-PCR analysis method is ΔΔCq, not ΔCq.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROBE-BASED QRT-PCR METHOD TO PROFILE IMMUNOLOGICAL GENE EXPRESSION IN BLOOD OF CAPTIVE BELUGA WHALES (DELPHINAPTERUS LEUCAS) Review round: 1 Reviewer: 3

Basic reporting: This was an interesting study on the immunology of captive belugas, providing preliminary data which may improve the health and welfare of captive cetaceans in the future. Although the authors have done a very good job as (presumably) non-native English speakers with the language of this paper, I would still suggest that they request a native English speaker to proof-read the text throughout and to name this person in the Acknowledgements section, according to PeerJ/ICMJE criteria. Overall, the English is good, however, occasional sentences, such as on lines 33-34, lines 65-68, line 54, etc. could be rewritten to ensure their meaning is clearer. In my opinion, the title is too long. I do not believe that anyone reading this paper needs the whole PCR procedure “spelling out”, surely this could be shortened by using qRT-PCR or RT-qPCR. The Introduction provides a good background to the subject. Literature is current and relevant. Perhaps it would be better to separate the Results and Discussion. This is the section order requested in the Instructions to Authors and it would appear quite straightforward to split this section. Although Figure 1 is clear and relevant, I am not sure that “note” is the correct name for the symbol used for sample C-I, perhaps this could be changed to a more obvious and more commonly-used symbol to avoid confusion. Experimental design: The research is within the scope of PeerJ. Research question: clear. It is stated that immunological data on cetaceans is limited and research is needed. Ethics: I have concerns about the “voluntary” aspect of the blood collection. While the authors state that international guidelines were followed (and I am sure that this is the case), these are not cited and I have been unable to find any such guidelines on the internet (I am sure they exist, but would like the authors should cite them directly). It is repeatedly stated in the results sections that beluga C showed reluctance to the “voluntary” blood draw and I feel that, when dealing with such sentient animals, this needs to be addressed and further explained by the authors. M&M: While it is clear from the raw data that 16 blood samples were taken in total (i.e. 4 from each animal), and not 16 from each beluga, this is not so obvious in the M&M section. Suggest rewriting slightly. I would also suggest stating how much blood was taken from each animal, presumably at least 1ml was taken to allow for the exact measurement of the 500 µl EDTA-blood to be preserved. Primer & probe design: Please state in the text that all gene accession numbers listed in Table 1 are from bottlenose dolphins. Results: Were any of the results statistically significant? Particularly the differences between the sick and healthy animals. I realise, however, that the low number of samples probably makes a robust statistical analysis impossible. Validity of the findings: Although 2 reference genes were used, they are not included in the results. Surely, if these genes were the control, they should at least be mentioned and not just included in the raw data. Conclusions: While I would agree that studies of this kind are important in an attempt to improve the health of captive cetaceans, I would suggest to the authors that they tone down their final conclusions and at least add a statement that their results are based on only two clinically symptomatic beluga and so may not be representative of the general cetacean population. Additional comments: Throughout: Shouldn’t RNAlater be written with a trademark symbol? Materials & Methods: Line 83: The beluga are in captivity, not in “human care”, which sounds rather strange. Line 100: Was gDNA contamination confirmed or excluded? Perhaps the absence of contamination was confirmed? Line 101: This is the first mention of cDNA so I would add the definition (complementary DNA) here rather than on line 118. Results & Discussion: Line 146: I would suggest that the opposite of highly transcribed is poorly (not lowly) transcribed. Lines 150-157: This text appears to repeat what is already shown in Figure 1. If there are any results in particular that the authors would like to highlight then I would do this here, but it doesn’t seem necessary to repeat all the information that is shown in the figure. Perhaps add the reference gene results here? Line 186: Were any tests for possible viruses carried out on Beluga C? It would be interesting to know what was the cause of the symptoms, plus different pathogens could have different effects on the immune system. Line 191: Perhaps add here that IL-4 is known to play an important role……….in humans, rather than just stating this as if the reference refers to cetacean research.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROBE-BASED QRT-PCR METHOD TO PROFILE IMMUNOLOGICAL GENE EXPRESSION IN BLOOD OF CAPTIVE BELUGA WHALES (DELPHINAPTERUS LEUCAS) Review round: 2 Reviewer: 1

Basic reporting: No comment Experimental design: No comment Validity of the findings: No comment Additional comments: All concerns have been addressed.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: A PROBE-BASED QRT-PCR METHOD TO PROFILE IMMUNOLOGICAL GENE EXPRESSION IN BLOOD OF CAPTIVE BELUGA WHALES (DELPHINAPTERUS LEUCAS) Review round: 2 Reviewer: 2

Basic reporting: No further comments. Experimental design: Thank you for adding the reference to animal handling guidelines. Could you please explain your comment in the response letter “When the animal showed reluctance to the voluntary blood draw, the trainers suppose some thing happens”? Perhaps you mean the trainers stopped the blood draw procedure and did some other training activity? It would be good to add this to the Materials and Methods section of the text if this is the case. The addition of Table 1 is a great help to the readers in understanding when blood samples were taken. Validity of the findings: No further comments. Additional comments: Lines 93-94: As the authors explained in their response letter, blood samples are taken from the fluke of the animal. As Beluga C was reported to have vesicles on the fluke perhaps it is important to mention that the area of blood draw may have influenced this animal’s “reluctance”. As such, I would suggest adding the following (or similar) to the blood sample method description “Sixteen blood samples……..were obtained via venipuncture of the fluke on a monthly basis…..” Line 94: Please change to “Within five minutes of 1-2 mL blood being collected, 500µl…..” Throughout: While RNAlater has been changed on line 96, ideally it should be corrected throughout the document. Line 207: As the authors stated in their responses that another research group is currently working on virus investigation, it would be informative for readers to mention here that this research is ongoing Table 1: please add “(years)” to the Age column Figure legends: For clarity ,I would add that the outliers refer to clinically symptomatic animals, e.g. “…..blood samples from clinically healthy (n=12) and symptomatic (n=4, outliers) belugas.”

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE SPANISH VERSION OF THE FATIGUE ASSESSMENT SCALE: RELIABILITY AND VALIDITY ASSESSMENT IN POSTPARTUM WOMEN Review round: 1 Reviewer: 1

Basic reporting: An article whose findings are important to those with closely related research interests. This manuscript present and discusses the validation of a scale to assess fatigue during the puerperium period. As author explain in the discussion section, this may be an apropiatte scale to be used in different time points besides the early postpartum period and helpful to assess fatigue evolution along all the postpartum As english is not my (reviewer) mother language , I would suggest an expert english language review. Experimental design: Reserch question is well designed and methodology used is adequate. Methods are well described and easy to understand. Hipotesis stated by authors is fulfilled. Validity of the findings: Analisys is well performed and is confirmatory. However as a suggestion to authors I would reccomend to obviate the exploratory analisis on item #4 and #10. It is already demonstrated and confirmed the saturation of these two items, and this exploratory analys may be considered as redundant. Additional comments: Manuscript is well written and easy to understand. Exploratory analysis for items #4 and #10 may be considered as redundant. Please consider reviewing This scale may be a good instrument to assess fatigue during postpartum period in Spain, and useful for international comparisons Minor comments: Line 122: authors use the acronym BSES-SF for the first time, please explain Line 135: it seems a wording mistake " closet o 0,95" , as i understand it is intended to be " close to 0,95". please confirm or correct

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE SPANISH VERSION OF THE FATIGUE ASSESSMENT SCALE: RELIABILITY AND VALIDITY ASSESSMENT IN POSTPARTUM WOMEN Review round: 1 Reviewer: 2

Basic reporting: The authors have developed a clear and rigorous scientific paper, with clear development and adequate language. The introduction is complete and allows the reader to be correctly placed in the context of the research. The references cited are useful and up-to-date. Complete graphic resources are provided and raw data is supplied. I suggest you take note of the following points about your figures and tables: Figure 1 - Write "Spanish" (as proper noun). Table 1 - Write "Table 1" (in English) instead of "Tabla 1" (Spanish). Table 2 - Add "#" symbol at ítem 4 and ítem 10. They can also do the same in the lines 142, 201, 210 and 217. Table 3 - Review spelling: "standard deviation". Experimental design: It is an original investigation with a clear purpose, which respects ethical codes, and whose method is simple and easily applicable. Validity of the findings: The conclusions are clearly related to the research approach. The validity of the results will allow an application of the scale in the daily consultations. Additional comments: I suggest you take note of the following format and spelling points: Abstract - Review spaces between numbers, words and fulls stops. Line 33 - Close quotation marks at the end of the sentence, Line 40 & 41 - I suggest you to link the expresions "fear and duration of labour" in one sentence. Line 74, 101 & 104 - Review spaces between words and brackets. Line 105 - Review spelling "maintain". Line 110 - Delete accent "version". Delete dead space at the end of the sentence. Line 111 & 112 - Review agreement between singular/plural in these sentences. Line 132 - Review spelling "and" ; "comparative". Line 145 - Review full stop of the sentence. Line 158 - Tabulate the beginning of the sentences. Line 159 - Delete innecesary full stop in the middle of the sentence. Line 162 & 164 - Review dead spaces. Line 179 - Write "molesta" and "preocupada" in italics (Spanish words) Line 203 - Full stop at the end of the sentence required. Line 216 - To maintain uniformity, I suggest you to to write "table 3" according to the other styles you use; "Table 3". Line 244 - Write "molesta" and "preocupada" instead of "molesto" and "preocupado" (maintain uniformity). Line 253 - Review space between full stops and words. Line 270 & 271 - Write "Model 1" and "Model 2" instead of "model 1" and "model 2" (maintain uniformity).

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE SPANISH VERSION OF THE FATIGUE ASSESSMENT SCALE: RELIABILITY AND VALIDITY ASSESSMENT IN POSTPARTUM WOMEN Review round: 1 Reviewer: 3

Basic reporting: The paper “The Spanish version of the Fatigue Assessment Scale: Reliability and validity assessment in postpartum women (#17227)” presents the cultural validation and initial psychometric properties of the FAS in postpartum women. The structure of the paper is adequate, contents are properly arranged in their sections, comments are not mixed with results, the discussion is well supported by the reported results, the methodology used is ade4quate for the aim of the study and the sample used has enough size. The paper is written in a way easy to understand although some English editing would be welcome. Some of the terms and expressions difficult to understand are pointed out bellow. Experimental design: The design is adequate for the aims of the study although some data analysis should be improved. Validity of the findings: The validity of results is supported by the large sample used. Methods are adequate, but perhaps incomplete, and regular procedures for questionnaire cultural adaptation have been followed. Sensitivity to change is still to be tested. Additional comments: To main issues are to be distinguished in the paper. One is the cultural adaptation into Spanish of the original FAS, the other is the existing differences in structure between versions. As for the first part, the paper is straightforward and results are really worth, specially taking into account the large sample used in the validation process. These findings are very useful in usual clinical practice and it is interesting to have a Spanish version of the instrument. Perhaps it would be interesting to have some additional data about the normative values obtained with this the sample, such as percentiles, in order to make it easier to know the meaning of the fatigue value obtained by a patient. As for the second part, there is nothing wrong in finding that the structure of an instrument might be different in other cultures. Nevertheless, I personally feel that the approach followed by the researchers is incomplete, but easy to work out. Reading the two items giving some trouble, they do remind other items used to measure anxiety and depression (like in the HADS scale). It is not easy to select proper items for a scale and the researchers are not the ones to be blamed for it. But some other methodological strategies are available to ensure that this two items are rather “noisy” different from the ones used by the authors. First. Estimated factor loadings are rather mild for a uni-dimensional scale. It could be the case that the authors have not considered the ordinal nature of Likert response scales and that the estimations method is not really the best one. I would recommend using MPlus defining ordinal metric for itmes and a better estimation method such as MLMV or WLSMV. Lisrel using the polychoric correlation matrix could also give good results. It would be interesting to mention in the paper the structure obtained by the original authors in English to compare. Second. The extraction method used in the exploratory factor analysis should be reported, and the obtained structure matrix to (even if the graph of rotated factor loadings is reported). Furthermore, it is not a good idea to use a Varimax (orthogonal) rotation when you are expecting items to be load in a single dimension. An oblique rotation would be more suitable, and the correlation between factors will give a hint on the real proximity of items in the rotated structure. The authors should keep in mind that they are not willing to improve the scale and hence they should not discard any item. At most, they should find better versions of the original items (¡and repeat the data gathering!). Third. The varimax solution will not give you the proper picture of the differential functioning of the two conflictive items. It is known that bi-polar dimensions (with a positive and negative end, and with items located at each one of the ends) tend to be treated as separated or independent dimensions by the rotation algorithms. The problem here is that the rotation procedure is not able to distinguish dimensions at 180º from those at 90º. This is a known analytical artifact, which should be thoroughly checked. Fourth. Given that the authors think that the negative wording of the offending items could be introducing a methodological bias they could use more sophisticated structures to test if this is so. For instance, setting free the correlation between error terms for items 4 and 10 (theta-10-4) would estimate the amount of relation of this two items due to wording, and subtract it from their relation with the underlying latent dimension. Other procedures as a bi-factor structure, a multy-trait multi-method, or a second order factor structure will also help understanding the structure and assessing the response bias due to negative wording. Fifth. The authors could always cross-validate the results splitting their sample and comparing both subsamples. From the results obtained (Line 225) It looks like there is not a big departure from the normal distribution (z=2.46) for the size of the sample used (n>850). If the authors are willing to compare mean values, they could safely use the t-test, since the distribution of possible sample mean values will be normal with such group sizes (no matter if the distribution of scores are not). Furthermore, it is more interesting to adjust for unequal group variabilities (when selecting the proper degrees of freedom for the t-test) than to prevent for deviations from normality (using the Man-Whitney U test). Z scores have the benefit that they give a feeling of the effect size for differences between groups, something much more difficult to do using the U. Some other minor comment follow. Line 58. The way the sentence is written makes you think that the fact that it is sort will make it more discriminant, or will give better psychometric properties. Please separate concepts. Line 58. Consider changing the expression “since it is brief, does” to “since it is brief, it does”. Line 62. Consider using “1=Never to 5=Always”. Line 63. Please modify the sentence. Only two items 4 and 10 are worded in a negative way, the other eight are positively worded but they ask for negative aspects or negative impacts on health i.e.: deterioration. Line 65. I suspect that the scores given to items 4 and 10 are not “negative scores” but rather “reverted” scores. See also Line 141. Lines 69. “reliability indices”. Only one reliability coefficient is reported. It should be singular. Line 105. Use “maintain”. Lines 109-110. The team ensuring semantic equivalence should have been the original English authors and not the developing team. Line 111. ¿How many interviews where carried out? Line 132: Use “and” for “ant”. Line 168. The name of local institutions is never translated. Although you may offer a translation in brackets. Line 182. How many cognitive interviews were carried out? It is customary to report a dispersion statistic (e.g.: standard deviation, interquartile range, etc.) accompanying the mean when average values are reported. See lines 176, 183 & 193. Line 195, 216. When particular figures and tables are addresses in the text, they should be mentioned in capital letters, i.e.: “Table 3”. Lines 199, 201, 259, 270. Please use the term “Goodness-of-fit statistics” or “Goodness-of-fit indices”, instead of “Indices of fit”, which is the standard for this kind of indices. Otherwise, use “fit indexes”. Consider rephrasing Line 199: “Goodness-of-fit statistics for the single dimension model containing all items (Model 1) were good.” Page 202. It is more frequent to use “factor loading” instead of “factor weight”, although it is not strictly incorrect. Line 208, please use chi-square (or 2) instead of c2. Furthermore, report the degrees of freedom after the chi-square value and using an equal sign: chi-square=2242.8, df=45, p<0.01. Line 229, 230. Use “FAS-e score” instead of “Score FAS-e”. Line 231. Use an equal sign between Spearman Rho and the reported value. Line 240. “owing to variance in method”. Consider using “due to a methodological bias”. Line 254. Consider using “no evidences have been provided” instead of “have data been provided”. Line 257. Consider using “support” for “attest to”. Line 265: it is difficult what the authors mean by “without leaving the common dimension”. Line 266. “problem of method variance”. Consider using “problem of variability due to method” or “variability due to method bias”. And modify Line 279 accordingly. Line 278: consider using “increase” instead of “exceed”. Line 281: Please review the following wording: “a sample of mothers to children aged 0 to 5 years old”. Line 290: I am not sure what the authors mean when using “diminish”. Perhaps they are intending to mean “discard”, but they should note that the kind of design they are using is not prospective and hence they may not infer if the FAS scoring (after birth) has any screening properties on self-efficacy (as related to vaginal delivery). Table 3. It should be mentioned that items 4 and 10 have been reverted before carrying all computations. Otherwise some results, such as the item-total correlation or the sign of the factor loading are difficult to understand. Table 3. Review translation for item 1. Table 3. Items should be numbered to make it easier the reference.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE SPANISH VERSION OF THE FATIGUE ASSESSMENT SCALE: RELIABILITY AND VALIDITY ASSESSMENT IN POSTPARTUM WOMEN Review round: 2 Reviewer: 1

Basic reporting: Authors have succsesfully reviewed and modified the draft accordint to comments and suggestions made in previpous review. Paper may be accepted now. No more comments to add Experimental design: No more comments to add Validity of the findings: No more comments to add Additional comments: No more comments to add

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE SPANISH VERSION OF THE FATIGUE ASSESSMENT SCALE: RELIABILITY AND VALIDITY ASSESSMENT IN POSTPARTUM WOMEN Review round: 2 Reviewer: 2

Basic reporting: The new version of the original paper “The Spanish version of the Fatigue Assessment Scale: Reliability and validity assessment in postpartum women (#17227)” has improved noticeably and the answers given by the authors to the reviewer suggestions are sound and clear. I would like to congratulate the authors for their efforts in improving their work. I would like to pose one minor suggestion. The term “group factor” is used to distinguish the content specific dimensions in the bi-factor solution from the overall factor/dimension (the later including all items). See lines 142, 229, 230, 239, etc. Although changing “group” by “grouping” could make it easier to understand the text it is usual, in the existing literature, to refer to such factors as “specific” factors as opposed to the “general” overall dimension (see Muthen & Muthen, 2010; Morin, Arens & Marsh, 2000; Reise, 2015). I would suggest using “specific” instead of “group”, especially when two samples have been used in the study, and it may induce readers to think about a multi-group cross-validation analysis using both samples. Please also correct it in Table 3 footnotes. Line 153 and footnote in Table 2. Use “goodness of fit index (CFI)” instead of “goodness fit index (CFI)” Line 185 and table 1. Use a dot as decimal placeholder instead of a comma. Line 213. Use “error terms” instead of “residues”. Avoid using “residuals”, since residuals in SEM models refer to the reproduced covariance matrix based on the model parameter estimates. Experimental design: No commnent Validity of the findings: No comment Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE SPANISH VERSION OF THE FATIGUE ASSESSMENT SCALE: RELIABILITY AND VALIDITY ASSESSMENT IN POSTPARTUM WOMEN Review round: 3 Reviewer: 1

Basic reporting: No comment Experimental design: No comment Validity of the findings: No comment Additional comments: No comment

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATING THE AGE OF HELICONIUS BUTTERFLIES FROM CALIBRATED PHOTOGRAPHS Review round: 1 Reviewer: 1

Basic reporting: The article entitled “Estimating the age of Heliconius butterflies from calibrated photographs”, by Dell’Aglio and colleagues, investigates whether the hue variation in the forewing dorsal red band of a Heliconius butterfly species is associated with age. The authors shot digital photos from several wings dissected from vouchers with known biological ages, calibrated the images to avoid bias related to light conditions and scale loss, projected the images on color space, and extracted information of the red-green opponency to use as data in the analyses. The results indicate that redness of red band and butterfly age are negatively correlated, suggesting that a red fading in the wings of Heliconius melpome butterflies is expected when adults get older. In general, the article is well written and have good quality data. The manuscript has a good discussion regarding the ecological and evolutionary implications of wing color changing in butterflies, and speculate about some studies that could be done in the future using the method presented in the paper, what I consider an inspiration for the scientific community. I really believe the method presented here represents an advance in the field. I do not have many criticisms regarding the paper. The only two concerns I have are: (1) why have been stated that categorical data was used in the analysis if only qualitative data has been used, and (2) why authors make the statement that analysis was based in human visual perception if all data came from a machine assisted visual perception? Experimental design: The experimental design is majorly ok, but the issues regarding (1) categorical vs. quantitative data analysis and (2) human vs. machine vision need further revision. IIn page 4 (lines 83-84), authors made the following statement: “we used an analysis based on human visual perception of wing colour to investigate the correlation with the age”. When I read that, I thought there would be an analysis based on human subjective categorization of red fading (see in figure 1), but to my surprise all analyses were done using quantitative data obtained from computer processed images. In the Material and Methods section, authors have justified why they consider that human perception had been used, but the truth is that "naked human eye" has not been used to identify the association between "redness" and age. Instead, machine processed photos, shot under standard conditions and with no noise produced by scale loss, were examined. Although all imagens had been taken under standard white light and processed considering only visible light spectrum, using RGB channels, which we humans can see, no human eye is capable of precisely identify a hue value of redness and at the same time exclude information from scale loss. Thus, the way I see, at least we should consider that a “machine assisted vision” was used. In addition, I did not find any analysis concerning the three categories of red (crimson, red, fade red) as described in lines 90-97. The only reference to the subjective age/red categories is made on figure 2, where authors discriminate the red categories of all wings measured using a*value in the same plot. Even so, all data used in the statistics and discussed in the results section came only from quantitative data (a* values). I think authors should consider that "real" human vision have not been used to identify the redness on the set of wings analyzed. Plus, I also recommend the authors to rethink about the necessity of include any reference to categorical data in the manuscript, since it not a major concern to answer the main question of the study (Is red fading of the wings associated with butterfly aging?). If authors choose to keep the references to the three classical age categories defined in the early 70's, I think it is needed to include another analysis comparing naked human eye efficiency in the identification of age classes using photos that have not been processed (scale loss edition). Validity of the findings: The manuscript have good quality data, and unequivocally indicates that red fading in the wings of Heliconius melpomene is associated with aging of adults. The issues regarding human vs. machine vision, and categorical vs. quantitative data do not affect the validity of the findings. Although similar results are expected in other species of Heliconius, the results presented here refer only to one species of the genus. Perhaps should be important to make this clear in the title of the manuscript. Additional comments: Although not discussed in the article, the results presented also raises important questions regarding the reliability of using only naked human eye and subjective categories based on color fading to identify the age of a butterfly. Figure 2 is very clear in showing that subjective red categories as a proxy for butterfly age are not good at all, since there is a massive overlap among categories. I think this issue could be included in the discussion (although not mandatory).

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATING THE AGE OF HELICONIUS BUTTERFLIES FROM CALIBRATED PHOTOGRAPHS Review round: 1 Reviewer: 2

Basic reporting: This study investigates whether the colour variation in the red ban exhibited by the butterfly Heliconius melpomene rosina is related to age. For this purpose the authors used digital photography and calibrated colour images of 55 individuals of this species whose age was known. The authors found a strong correlation between hue and age, according to human visual perception, so that younger individuals display a more intense red band (crimson) whereas older individuals present a faded red. The interpretation of these results is framed within literature on age-related colour fading, mate choice and possible relation with predator avoidance. The figures are relevant for the paper, and the one related to the results is self-explanatory. The references cited are appropriate for the topic of the study. The paper is well written in general, although I have some suggestions for improvement. Experimental design: I appreciate the information provided, considering that many of these assumptions (i.e. ‘redness’ is related to age) spread despite not having been tested. However, I think that the justification of your study needs to be reformulated. You say “Given the importance of red patterns in mate choice, we here investigated the influence of butterfly age on red colouration. We used an analysis based on human visual perception…”. If the rationale for your study is related to mate choice and, for example, whether or not age (through “redness”) could influence mate choice, using human visual perception is a strange choice. According to L164., vision has been studied in related species, so data should be available to give “butterfly values” to the differences in redness among individuals, according to your calibrated images. With the current knowledge on avian visual models, you could also check whether birds (potential predators) detect that variation in “redness”. However, with your choice of human perception these questions cannot be addressed. Thus, you need to re-formulate the justification of the study, placing more emphasis, for example, on why it would be useful for researchers to have a phenotypic indication of age. This would also make more ‘acceptable’ to include speculation in the discussion section, and would be a first step, as you imply, to address the relationship between colouration and age. L114: The line about the statistical analysis is confusing. You state that you checked for a correlation between two variables using a linear regression, and then report a ‘c’ value in line 126… is this ‘r’ (from a correlation analysis)? A regression analysis is used when you want to predict a Y-value for a varying X-value, usually which is controlled by the researcher. I agree with your statement saying: “to test for a correlation….”, as this is what you can do to see to what extent, if at all, age and ‘redness’ are related. Also, you need to mention whether you did a parametric (Pearson) or a non-parametric (Spearman) correlation? Finally, you report no differences in colour between males and females with a t-test (L124). I would suggest you use a one-way ANOVA instead (I am assuming your colour data are normally distributed)? L125-127: Probably the most important line in your paper, yet not clear in terms of the statistical approach. Is ‘c’ the correlation coefficient or an r2? If the former, that means you did a correlation analysis (which is, in my opinion, the right thing to do) rather than a linear regression. Much more accuracy about what you did is needed in the results section, even if I can see from the figure what the results are. Validity of the findings: RESULTS ---> L125-127: Probably the most important line in your paper, yet not clear in terms of the statistical approach. Is ‘c’ the correlation coefficient or an r2? If the former, that means you did a correlation analysis (which is, in my opinion, the right thing to do) rather than a linear regression. Much more accuracy about what you did is needed in the results section, even if I can see from the figure what the results are. Additional comments: See below a list of more specific comments. L50: Please add references to support this statement! L51: change ‘particular’ to ‘particularly’; add an ‘s’ at the end of ‘exhibit’, as here you are treating Heliconius as a genus (i.e., single) L52: replace “that they are toxic” with “of toxicity”; also, insert a comma (,) after the brackets L53: something is wrong here, I think? Do you mean “and find and choose mates based on the same colour signals that are involved in predator deterrence”? If I understand correctly, you want to say that the same colour signals are used to attract mates and repel predators… Right? L58: ‘Melinaea’ should not start with capital letters L62-63: change “for variation in colour” to “of colour variation” L64: change ‘are’ to ‘is’, as you are referring to ‘gene expression’ (singular) L73: insert a comma (,) after ‘genetically’ L78: … and predator avoidance(?) L80: delete ‘the’ at the end of this line L79-83. Consider splitting this line in two L93: replace ‘captured’ with ‘taken’ L146: add ‘in’ before ‘which’ L162: remove the ‘s’ at the end of ‘shifts’ L163: add a comma (,) after sensitivity and replace ‘and allows’ with ‘allowing’

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATING THE AGE OF HELICONIUS BUTTERFLIES FROM CALIBRATED PHOTOGRAPHS Review round: 1 Reviewer: 3

Basic reporting: see the 'General comments for the author' below Experimental design: see the 'General comments for the author' below Validity of the findings: see the 'General comments for the author' below Additional comments: The work reported here investigates whether the hue variation in the forewing dorsal red band corresponds to the age of the butterflies. Tested with 55 digital wing images, it is claimed that red colour can be used as indicator for age based on the strong (negative) correlation between age and a* values. As evident in Fig 2., all Crimson wings were <10 days and faded reds were about 30+ days old. So these two subjective colour labels could only put estimated age in these bands. The labelling of ‘red’ seems to cover a wide range (from 1-40 days) and hence not at all useful for age prediction. On the other hand, a* values showed stronger correlation with age but this was not thoroughly investigated. Is this correlation value due to the distribution of the limited data in two fairly distinct clusters? There are very few samples in the age range 12<age<30. Hence the regression line is unlikely to be a good predictor of age. This could easily be tested by setting aside a subset of samples (as test set) from the regression analysis and the regression line obtained from the remaining data be then used to predict the age of the test set samples. Not being a bio-scientist/ecologist myself, I am not sure what sort of precision is generally expected or necessary in such age prediction exercise. But my personal feeling is that the proposed approach is not adequate in ‘estimating the age’ precisely enough. [I was curious to know if the two wings of the butterflies demonstrated similar colour values since they are the same age]. The authors may explore other colour representations as the discriminatory feature and preferably use a much larger data set to see if better age estimation is possible. Alternatively, if a rough estimation is good enough, this should be explicitly stated and then demonstrated how accurate are the age estimations under this definition. The paper is generally well structured, language easy to read. Some (very few) typos still managed to get in which can easily be fixed.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATING THE AGE OF HELICONIUS BUTTERFLIES FROM CALIBRATED PHOTOGRAPHS Review round: 2 Reviewer: 1

Basic reporting: I am happy with how the authors have addressed my comments from a previous version. Experimental design: The authors have done the suggested changes to the manuscript in order to make their aims clear. Validity of the findings: Ok Additional comments: Thanks for having carefully addressed my previous comments. I have nothing left to add.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: IDENTIFICATION OF REFERENCE GENES IN BLOOD BEFORE AND AFTER ENTERING THE PLATEAU FOR SYBR GREEN RT-QPCR STUDIES Review round: 1 Reviewer: 1

Basic reporting: ─ With respect to the volunteers the clinical and experimental information should be expanded and some basic characteristics should be provided, e.g. age, sex, time spend in the plain before moving and time spent on the plateau before blood was drawn. Did any of the six volunteers they show clinical signs of hypoxia at the time of examination? ─ There are some minor typos that should be corrected, e.g. line 86 “qRCR RT kit”, line 129 “quantitative quantification method”. Experimental design: ─ The limited number of volunteers the experiments are performed on should be discussed critically. Validity of the findings: ─ Please report the number of replicates the measurements of the different housekeeping genes was performed on. ─ What do the authors think is the reason that these housekeepers reliability is different depending on the evaluated altitude? Additional comments: In the presented study, the authors assessed the stability of eight different human housekeeping genes in standard SYBR Green RT-qPCR under normoxic and hypoxic conditions. Different algorithms (geNorm, NormFinder, BestKeeper and RefFinder) to evaluate the stability were applied. In their study, they identified RPL13A as most reliable housekeeping gene for evaluation of gene expression in normoxic and hypoxic conditions.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: IDENTIFICATION OF REFERENCE GENES IN BLOOD BEFORE AND AFTER ENTERING THE PLATEAU FOR SYBR GREEN RT-QPCR STUDIES Review round: 1 Reviewer: 2

Basic reporting: English proficiency requires improvement throughout the paper. Examples: -line 44: In the abstract: "Interestingly, RPL13A was selected as the ideal reference gene to normalize the target gene expression in human blood before and after moving to the plateau." I suggest changing the selected to identified, or found ... - line 52: What do you mean by "Hypoxia is a major geographic feature of high-altitude regions", the word geographic does not fit here, may replace with biological? - line 53: the following sentence is not clear: "In hypoxic environment, specific genetic programs and molecular mechanisms initiate various genetic events." - line 54: "hypoxia-inducible factor (HIF) oxygen-signaling pathway" do you mean alpha not oxygen? - line 154: what is (Vn / Vn + 1 = 0.15)??? please explain. -line 172: "is based on the coefficient variance (CV) and standard deviation (SD)" of what? please complete the sentence. - line 203: what do you mean by: "at both stages of entering the plateau"? -Figure 3: A, B , and C, why it is connecting line, these are separate values, are they related to each other? -table 5, 6, what do you mean by whole stage and whole stages, is it plain and plateau combined? Experimental design: suggestions to authors: please provide more details related to the study design, how many samples were collected from each individual and when, how many days after they were moved to the plateau. Validity of the findings: Overall, the idea of the study is interesting and might be useful for future researchers in the field of hypoxia and the selection of appropriate housekeeping genes. Yet, the authors do not explain the variability in their findings related to the different ranking scores among the different validation programs they used. This limitation of the study should be addressed in their discussion. Also further discussion should focus on the limitation and reliability of each of the used programs (geNorm, NormFinder and BestKeeper). Would it be feasible for researchers to validate their finding in an experiment where cells are placed in hypoxic chamber and the same reference genes are assessed? Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: IDENTIFICATION OF REFERENCE GENES IN BLOOD BEFORE AND AFTER ENTERING THE PLATEAU FOR SYBR GREEN RT-QPCR STUDIES Review round: 2 Reviewer: 1

Basic reporting: The English - including the newly added parts - would still benefit from improvements. Experimental design: no comment Validity of the findings: no comment Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: IDENTIFICATION OF REFERENCE GENES IN BLOOD BEFORE AND AFTER ENTERING THE PLATEAU FOR SYBR GREEN RT-QPCR STUDIES Review round: 2 Reviewer: 2

Basic reporting: The manuscript is still unclear, professional English editing is critical and required for this manuscript before being accepted. examples: Line 78: what do you mean by “Gene sequences were deposited in the NCBI database under GenBank accession numbers indicated in Table 1” Line 96: “correlation coefficient (R2 ) of each candidate reference gene were calculated for (Table 2)” calculated for what??? Line 100: “The geNorm (Vandesompele et al. 2002) program is a measurement of gene expression stability 101 (M)” the program is not a measurement. Line 113: add “in subjects” after the word reference genes Line 138: need to explain in the text what the formula means, what V is. Line 181: “However, there are few studies on the evaluation of optimal reference gene(s) 182 between low- and high-altitude conditions” this sentence is redundant. Experimental design: details are still lacking: RNA samples and cDNA synthesis: this section still lacks a lot of details, such as how samples were processed, were the mononuclear cells isolated? What method as used? Or did they pelleted the whole blood? How did they extract the RNA There is no positive control to confirm that hypoxia has occurred in the blood of the subjects after 3 days of being in the plateau. A measurement of hypoxia related gene is needed, and/or validation of these data in experimental hypoxic conditions Validity of the findings: fig. 1: why there is no band for TBP and very faint band for SDHA? and there was no mention in the text regarding this Table 5: what is CP? Discussion: it does not address the limitation of the study, the tools that are used, and the significance of their finding. It is more or less repetition of the results. Additional comments: Since the experiments related to validating these findings in hypoxic chamber is underway, these data should be feasible to add.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATED EFFECTS OF IMPLEMENTING AN OPEN ACCESS POLICY FOR GRANTEES AT A PRIVATE FOUNDATION Review round: 1 Reviewer: 1

Basic reporting: 1. The methodology is not very detailed (insufficient detail for replication), with some errors (see below). 2. A structured abstract would be preferable to make it clear this is considered research. 3. It would be helpful to have a basic description of the number of grants and the size of grants. The total amount of funding the GBMF provides annually should be provided, to provide a frame of reference. 4. “OA journals provide access to all content immediately online.” Most definitions of open access also include Creative Commons licenses, but they are not addressed until the last paragraph of the article and their benefits are not mentioned. If the authors are not considering CC as part of this evaluation, they should define up front that the type of OA being analyzed refers to free content only. 5. “We then estimated the annual cost for making these articles OA by multiplying by a low ($1500) and high ($3500) APC estimate (Solomon 2013).” This reference is not included in the reference list (and should be) so it wasn’t possible to verify that the estimates are reasonable and appropriate for the journals in the study. 6. “Instead, they are often more concerned with publishing in the most relevant, highest impact journal for their field (Priem 2013).” Priem 2013 doesn’t appear to say this—it is a thought piece about the future of how researchers will disseminate their work and indicates that researchers are already changing in how they do so. Is the reference wrong? 7. “To explore potential impacts of an open access policy at GBMF, we analyzed 2650 publications produced by GBMF grantees between 2001 and 2017” Did the choice of journal change over time? 8. “Figure 1. Breakdown of journal type (open, hybrid, or closed) used by GBMF grantees for the 2650 articles (left) published in 573 journals (right).” This notation of journal vs article is opposite what the figure shows. Which is correct? 9. Dataset: would be more transparent to review if Post Print Policy indicated Y for authors are allowed to publish and N for they are not allowed to (rather than 1 or 0). Experimental design: 1. The novelty and importance of such an analysis is not addressed. Have any other foundations published their analysis? Why is it important to share this information? 2. The design is not strong; the authors have no way of knowing how complete the data are; certainly the data in 2015 imply that the rest of the years are underestimated. The estimates vary 10-fold. The estimates for APC charges are simplistic and the reference for them is not listed. 3. The authors don’t actually provide the calculus of weighing pros and cons besides concluding that gold OA would be expensive (apparently – they don’t actually say that they concluded it would be too expensive for the foundation). The fact that gold OA would cost more than green OA is apparent without doing the analysis. Validity of the findings: 1. The authors do not know the number of papers actually published (the statement “We identified more than 800 publications in 2015 (compared to approximately 300 and 500 in the years before and after, respectively). In 2015, the newly funded Data-Driven Discovery Initiative in the Science Program began systematically collecting data from its grantees about publications, resulting in the large spike.” suggests that the articles in other years are substantially undercounted. 2. The gold PA cost estimates have a 10-fold range and are based on assumptions rather than actual charges; they acknowledge that the actual amounts may be lower or higher. 3. Factors such as institutional membership that would reduce expenses, or OA journals that do not charge APCs, are not taken into account. 4. The authors do not estimate the value of OA, and in fact only mention the benefits of OA in the introduction. 5. They don’t consider Creative Commons licenses. 6. Title, “Expected effects of an open access policy at a private foundation” This study does not address effects (including impact of the research if it is made freely available after a year, immediately, or never), only costs and journals that authors could publish in. Perhaps “Estimated costs of implementing an open access policy at a private foundation” 7. The authors state, ““OA offers a number of benefits including increased citations counts (Gargouri et al. 2010), accessibility for building research capacity in developing countries (Chan et al. 2005), enhanced visibility of research (Tennant et al. 2016), and decreased financial pressure on academic and research libraries (McGuian and Russell 2008).” They state, “In an effort to increase access to the research results it funds” but which specific benefits were of greatest importance for the GBMF? How does their new policy address these benefits? 8. Why did the GBMF feel it was important to let authors publish where they normally would rather than encouraging them to publish in OA journals 9. “grantee’s preferred journal” – it’s unknown what their preferred journal was; this was the journal they published in. 10. “GBMF plans to revisit the efficacy and impact over time for the newly implemented policy. There are several variables that can be altered to potentially strengthen the OA policy. These may include (1) restricting embargoes on archiving OA versions to six months (compared to 12 months); (2) requiring CC-BY licenses for all publications; (3) setting aside funds for covering grantees’ OA fees (independent of their grant funds); (4) mandating particular repositories for OA archiving; or (5) expanding the policy to include outputs other than peer-reviewed journals (e.g., data, software, books, etc.).” It is surprising to see this statement after the analysis. Why was the decision to use the green approach made and what factors would make GBMF rethink that decision? 11. “GBMF funds research in basic science, environmental conservation, patient care improvements, and preservation of the San Francisco Bay Area.” “Moore Foundation grantees tend to publish in hybrid journals (74% of journals; 72% of articles), with open (16% of journals; 17% of articles) and closed (10% of journals; 11% of articles) journals less represented.” The ratios of gold to green journals vary substantially by field (see Björk et al., 2010, doi: 10.1371/ journal.pone.0011273.g004 ). It would be useful to see whether the calculations vary substantially by field. 12. Cost estimates don’t take into account memberships (eg, In “Cambridge RCUK Block Grant spend for 2016-2017” https://unlockingresearch.blog.lib.cam.ac.uk/?p=1463 : “The average article processing charge was £1850 – this is significantly less than the £2008 average last year, reflecting the value of memberships”) Additional comments: This paper provides the assessment conducted by a foundation in determining what OA policy to implement, and you should be lauded for being more transparent about your decision making than is often the case. However, the paper has a number of limitations. The limitations are in part due to the challenges of finding all the publications supported by GBMF and taking into account the actual APCs instead of the estimated APCs. However, if you don’t provide the caveats and frame the discussion appropriately, another foundation might look at this and make the same decision without digging deeper, which would be unfortunate. The decision-making process presumably involves weighing pros and cons, not simply (apparently) "that's a lot of money, we can't afford that." Therefore, while the transparency is laudable, the analysis is fairly simplistic and it doesn't really rise to the level of research. While a formal cost-benefit analysis is presumably more than you are willing to undertake, what proportion of grants would have to not be funded to make up for the expense of funding OA? This sort of estimate would be useful.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATED EFFECTS OF IMPLEMENTING AN OPEN ACCESS POLICY FOR GRANTEES AT A PRIVATE FOUNDATION Review round: 1 Reviewer: 2

Basic reporting: This paper presents an initiative of the private funding company for changes in their policy related to the open access to the articles written by the grantees. The study involves “more than 2000” publication in “more than 500 journals”, and approximate cost of OA, according two suggested OA policies, has been calculated and discussed. There are wrong definition on gold and green OA present: “Gold Open Access” journals which are considered only as journals with APC business model. It is well known that many gold OA journals do not charge authors. Also "Green Open Access" is not related with the journals but with authors who deposit their works (not only journal articles) into different institutional, subject based or other type of OA repositories. In the focus of the study are possible OA policies, but it is not clear why the funder wants to mandate OA. Benefits of OA for funders and researchers are not discussed. Funders have encouraged or mandated OA because enhanced visibility and discoverability increases the return of their investment, making the results of the funded research more widely available. Funder’s OA policy ensures the results of the research they fund can be read and used by anyone, including industry and society. Also, funder adopting OA policy avoids funding duplicative research, creates transparency, and encourages greater interaction with results of funded research. There are more benefits for researchers/grantees too. None of these are mentioned or discussed. There is only the phrase “potential implications” used in the discussion, but none “potential implication” is mentioned or discussed. The manuscript was written in clear and professional English. Literature references, although not sufficient, are context provided. The missing literature includes mostly papers discussing OA mandates of other big funders. The manuscript is well structured and two figures are provided. Raw data is shared partially. Data on journals are included, but data on articles are missing. The title is somehow misleading since no “expected effects of an open access” are discussed in the manuscript Experimental design: Research questions are defined as an interest “in understanding (1) whether grantees would be restricted from publishing in their preferred journals, and (2) the financial ramifications of a policy advocating for and funding Gold OA when available.” Authors are not mentioning possible influences of the funders and the present research assessment criteria on the “preferred journals”, which could contribute to the study. There is also kind of confusion with the years included in the study: 50 “In an effort to increase access to the research results it funds, The Gordon and Betty Moore Foundation (GBMF) began considering the implementation of an open access policy all publications produced by its grantees in 2016.” 65 “To explore potential impacts of an open access policy at GBMF, we analyzed 2650 publications produced by GBMF grantees between 2001 and 2017.” 76 “We also calculated an annual estimate of Gold OA costs for 2009-2016.” Also, APC price ranges are not used consistently: 39 …on the order of $1000-3500 USD – USD needs to be deleted later on, estimating the annual cost for making articles OA, authors are multiplying by a low ($1500) and high ($3500) – is it “low” $1000 or $1500?? Limitations of the study are not present. Validity of the findings: Other minor changes include: 37 …OA are available publicly online. 77 …by 0.89 – what kind of data from Sherpa/Romeo was used to calculate this percent? 88 “Although we have no information on the percentage of those articles published in hybrid journals that were published Gold OA, we assume the percentage is low since this was not a requirement for grantees.” – information on OA articles by grantees could be easily retrieved, there is no need to estimate 102 “Based on these proposed policies, we calculated that the percentage of articles not compatible with Policy 1 is 0.7 - that is, 99.3% of articles would have been compatible with Policy 1.” – please leave not compatible, OR compatible part of the sentence. 104 two times “were” 105 Annual reviews are “explained” but there is no single reason present which can explain why they are close 120 “since these are the years where we have data for more than 100 articles.” – what kind of criteria is that?? 121 in the Results expressions like “more than 800”, “approximately 300” and “approximately 500” cannot be used – please state accurate numbers Additional comments: An interesting topic was presented that has been somehow. superficially elaborated.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATED EFFECTS OF IMPLEMENTING AN OPEN ACCESS POLICY FOR GRANTEES AT A PRIVATE FOUNDATION Review round: 2 Reviewer: 1

Basic reporting: The manuscript is substantially better reported and nearly all the original concerns have been met. There are a few points that should be clarified, below. 1. Title is better but still could be more clear, eg, "Estimated costs of implementing an open access publication policy for grantees of a private foundation" 2. Background: The title was revised but the Background is still overly broad. Perhaps: "The Gordon and Betty Moore Foundation (GBMF) was interested in understanding the potential costs of a policy requiring that grantees publish their peer-reviewed research in open access journals." 3. "Anecdotal evidence suggests most authors do not select journals based on their level of openness or their policies around self-archiving. Instead, they are often more concerned with publishing in, instead choosing the most relevant, highest impact journal for their field.": Without a reference the evidence is anecdotal, therefore the language should be more equivocal, eg, "anecdotal evidence suggests that authors may not select journals based on..." 4. "We also determined whether the journal allowed authors to archive post-prints (green OA)": "Post-prints" is not a commonly used term--"archive accepted or published manuscripts in a publicly accessible repository" would be more clear. 5. "Based on the data reported here, the OA policies considered by GBMF are unlikely to have significantly impacted the grantees’ choice of journal due to lack of compliance.": "due to lack of compliance" is unclear -- maybe "would not have substantially altered the journals in which the grantees published their work" ? Experimental design: The authors now adequately describe the limitations of the design in the Methods and Discussion. Validity of the findings: The authors have addressed issues regarding impact and novelty, and corrected previous errors. Conclusions are appropriately circumspect. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ESTIMATED EFFECTS OF IMPLEMENTING AN OPEN ACCESS POLICY FOR GRANTEES AT A PRIVATE FOUNDATION Review round: 2 Reviewer: 2

Basic reporting: no comment Experimental design: BEFORE: “Although we have no information on the percentage of those articles published in hybrid journals that were published Gold OA, we assume the percentage is low since this was not a requirement for grantees.” – information on OA articles by grantees could be easily retrieved, there is no need to estimate REVISED: „Although we have no information on the percentage of those articles published in hybrid journals that were published gold OA, there is evidence that without funder mandates, the percentage of articles that are published OA is low (Tennant et al. 2016).“ Authors' response: This work is beyond the scope of the project and would require a significant amount of additional research and work. It was not part of our original research questions. The authors are not aware of where this information can be “easily retrieved” in bulk. Could not agree that this is „beyond the scope of the project“. It is very important from the funder perspective to know exactly what percentage of articles are already OA (without additional funder's investments in OA), and before the implementation of the OA policy (mandate). Since the hybrid journals are the most represented group (74%) such an evidence should be the starting point of this study. There is the way to do it „in bulk“ for the articles in HTML and PDF formats. If authors cannot do it „in bulk“ for various reasons, the data for the last two or three years (2014-2016) could be checked manually and provided. „We were also interested in whether a grantee’s journal choice would be impacted by two possible OA policies being considered. Policy 1 would require OA within 12 months of publication, either via the green or gold route. Grantees could comply by either publishing in open journals, by publishing OA in a hybrid journal (gold OA), or by publishing non-OA and self-archiving within 12 months (green OA). This policy would preclude grantees from publishing in journals that are closed and restrict self-archiving for longer than 12 months. Policy 2 would require immediate access at the time of publication (gold OA). Grantees could comply by either publishing in open journals, or by publishing OA in a hybrid journal. This excludes journals that are closed and/or restrict self-archiving at the time of publication.” - The only difference between Policy 1 and 2 is the 12 month embargo period which reflects only on articles from the closed journals and the green OA (self-archiving) share. Green OA is free of charge, and there are no differences in costs for OA between Policy 1 and 2. Since there are different journal policies on self-archiving, some articles from the closed journals will be allowed for self-archiving immediately at the time of publication (or even before!), and some not. Both policies include gold and green route to OA. This is not clear from the descriptions of the Policies. Validity of the findings: Abstract: “Based in part on this study, GBMF has implemented a new open access policy that requires grantees make peer-reviewed publications fully available within 12 months. – the authors did not consider the usage rights (Creative Commons licenses), although it would be necessary for such an analysis, and this should be corrected in „available for reading“. The articles that would not have been compatible with Policy 1 were exclusively from the family of Annual Reviews. These journals are not typical publications since they do not publish original research; they are review series in specific disciplines in science and social science. – there are Annual reviews also in other disciplines (Annual Review of Applied Linguistics, Annual Review of Biomedical Sciences, Annual Review of Clinical Psychology, Annual Review of Energy, Annual Review of Genetics, Annual Review of Immunology, Annual Review of Neuroscience, Annual Review of Nutrition, Annual Review of Pharmacology and ToxicologyAnnual Review of Public Health, etc.) „The estimated maximum cost for implementing Policy 2 (i.e., gold OA for all publications) ranged from $401,835 in 2008 to $2,538,725 in 2015. (Figure 2).“ – the estimated cost is valid for both policies since the only difference is embargo (and green OA respectively). „For example, we identified 814 publications in 2015 (compared to 328 and 501 in the years before and after, respectively). In 2015, the newly funded Data-Driven Discovery Initiative (https://www.moore.org/initiative-strategy-detail?initiativeId=data-driven-discovery) in the Science Program began systematically collecting data from its grantees about publications, resulting in the large spike.“ – the given explanation does not explain the lower number of publications in 2016 when Data Driven Discovery Initiative was still in place. “Our estimates of number of publications per year (and therefore estimated maximum APC costs) rely on either the grantee self-reporting to the foundation, or the grantee including GBMF in manuscript acknowledgements that can be harvested by Crossref. Our range of costs is therefore likely an underestimate for earlier years that have lower publication counts due to the increased difficulty in tracking down data from those years.” – It is not clear are the difficulties in tracking down data for the most recent years solved or not, and are the numbers of articles provided for the most recent years in the analysis accurate. For example, it is not explained why the number of articles in 2016 is significantly lower, compared with 2015. Additional comments: This is improved version of the article, but the general impression is that experimental design has not improved significantly. The number of articles per year is not complete, and even though the authors state limitations in the collection of articles, it is unclear why the data for the latest years (except 2015) are not more reliable. The data on the share of articles already available in the open access is missing, and I strongly suggest the authors to add this, if not in this one, then in their future research. The authors' opinion is that usage rights and licensing are beyond the scope of the study, but this is an important issue for funders and should be considered as a part of the implemented policy. Open access is more complex than "gold" and "green" route.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ANANKE: TEMPORAL CLUSTERING REVEALS ECOLOGICAL DYNAMICS OF MICROBIAL COMMUNITIES Review round: 1 Reviewer: 1

Basic reporting: The paper was well written and the figures were useful. No problems with basic reporting, but see comment about Preheim et al. 2013 in the general comments below which I think should be cited. Experimental design: The research identifies a problem, and definitely addresses it. They perform some analyses based on simulated data that is nice and well described. Some of the short comings of the algorithm and future directions are clearly stated. Validity of the findings: The conclusions are one of the strengths of the paper where they do a great job of recapping the positives and negatives of the approach. Additional comments: Hall et al. Ananke Review The contribution by Hall et al. provides an informative description of the bioinformatic approach they have developed. This software enables clustering of exact sequences across time-series based on their ‘dynamic similarity’. The paper is very well written and the software is a timely complement to OTU based approaches. The approach may be even more useful to help analyze datasets partitioned with finer resolution such as DADA2 and Minimum Entropy Decomposition, as more investigators transition their analyses to these methods. In addition to Ananke, the Shiny-based visualization is very cool, useful, and is key to assessing Ananke’s ability to discriminate temporally similar sequences. Below I highlight some aspects that I find should be addressed, roughly, in order of importance. An important point, in general, is that the authors seemed to leave out what would seem to be an important reference to work that is similar. They need to cite and discuss a paper by Preheim et al. 2013, “Distribution-Based Clustering: Using Ecology To Refine the Operational Taxonomic Unit.” The Preheim et al. software uses both sequence similarity and distribution of their abundance across samples, including discriminating above 97%, similarity as Ananke does. The authors need to explicitly address the similarities and dissimilarities between these two approaches, which seem quite similar. Ananke is different in that it does not require sequence similarities, so it can find co-occurrence clusters between non-closely-related taxa. Which may or may not be considered an advantage. Line 300: Ananke does not perform consistently across datasets of variable lengths, and they speculate that it might differ based on sequencing depth and diversity. They suggest that the user needs to go use the visualization tools provided (which is very nice and works great) and set their cutoff thresholds based on detailed assessment of the data. However, I find this requirement at odds with their statement in the title of section at Line 103: “Unsupervised clustering of time-series distances”. I recommend omitting “Unsupervised”, here because even if they calculation is unsupervised (like OTU clustering), the assessment and analysis, does not seem to be so. Section starting at Line 178: Title seems to disagree with my understanding of Line 95 and others. The authors found that more data meant less similarity to the “ground truth” not that data sets were “accurately clustered”. Authors state that the algorithm has a difficult time on taxa that are normally distributed with high or low variance, and on datasets with many time-frames, I appreciate their forthrightness. Though I recommend that the authors reflect this in the title of the section: Perhaps, “Assessing accuracy of Ananke with simulated datasets” Line 319: I recommend the authors expand the point to include it’s ability to complement non-OTU based approaches like DADA2 (Callahan et al. ), MED (Eren et al.), and the Tikhonov method. Line 214: Ananke cannot be used on taxa that are detected in less than 15 or 20% of samples. Given the importance of the rare biosphere, of course this is a drawback. I think that the authors should state this, but can also point out that these taxa will be found in the “noise bin”? It is a good example of why Ananke should be considered as a complementary approach to other methods, as stated in the conclusions. Line 250: Many of these analyses may be performed on more powerful computers than a standard desktop. How is Ananke expected to perform on these rare taxa if one doesn’t filter them out? I imagine Ananke will perform well on some of these periodic taxa if they become very abundant on a few days, and not so well if they are just always around the limit of detection (very stochastic). General comment: I think that the software would be more popular, accessible, and more cited if the authors wrote some additional guidance in the github page to give users examples of how Ananke could be used in parallel with other extremely common pipelines (QIIME and mothur) or (where I think it will be most useful) very quickly up-and-coming methods (MED and DADA2). This wouldn’t need to be in depth, just some general guidance. Also, the software was very easy to install with root privileges, great job! Might consider guidance for users without root privileges. Also might consider providing a test dataset for testing if installation was fully successful and working as expected. Line 53: Probably want to cite earlier work, e.g., Eren et al. 2013. Line 57: The Callahan reference is now published in Nature Methods, 2016. Line 63: Is Ananke an acronym for anything? Line 230-232: Authors should keep in mind that some taxa may have multiple copies of 16S and sometimes the sequences can vary (I’m not sure about E. coli). This could be a type of positive control for their, or other investigators, analyses. Section beginning at 298: The influence of the user’s selection of epsilon values on the results is very clear and appreciated. Line 311 Aren’t Actinomycetales and Acidimicrobiales relatively closely related? Perhaps an indication of the sequence similarity of these taxa would be useful.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ANANKE: TEMPORAL CLUSTERING REVEALS ECOLOGICAL DYNAMICS OF MICROBIAL COMMUNITIES Review round: 1 Reviewer: 2

Basic reporting: see General comments for the author Experimental design: see General comments for the author Validity of the findings: see General comments for the author Additional comments: Hall et al. developed a software Ananke to cluster the time-series microbiome data. There are several major issues in this study. The processing pipeline in Ananke consists of: 1) Tabulate unique sequences 2) Filter out sequences with low counts 3) Calculate short time-series distance and 4) Cluster using DBSCAN. Unique sequences in the microbiome data can be easily calculated using 100% identity OTUs in the software QIIME or Mothur. Calculation of unique sequences was not performed in the previous studies, which was not due to the technical or software problems. The aim of the operational taxonomic units (OTUs) clustering is not only the data simplification but also minimizing the influence of sequencing errors. The sequencing errors occur in all the sequencing platforms, particularly higher in 454 pyrosequencing. Using unique sequences rather than OTUs would make the analysis intolerance of sequencing errors. The 97% sequence similarity as OTU clustering cutoff was widely used in the 16S rRNA gene analysis, largely representing the community of the gut microbiome at the species level. 98% and 99% OTUs were used in the oral and skin microbiome studies as well. Unique sequences or 100% OTUs do not necessarily increase the resolution of the data analysis in the hypervariable regions of the 16S rRNA gene analysis. In addition, Ananke takes the fasta file as input, ignoring the quality information (scores) of the sequencing data. As shown on the simulated time-series data sets in the study, Ananke performed well with as few as ten time points. However, the data noise or error significantly affects Ananke over a large number of time points. Ananke has also the trouble in clustering normally distributed time-series data. The analysis based on unique sequences significantly increased the computational memory requirement and the running time. The software performance, RAM usage and running time, was not evaluated in the study. Thus all the above suggest that the rationale of the 1st step in the method does not make sense. It’s not a good idea to analyze the unique sequences in the microbiome study. To make the method feasible, in the 2nd step, Ananke minimized the number of unique sequences by filtering out the ones with low abundance. A large proportion of data were missed in the downstream analysis, ignoring numerous species/strains in the microbiota and affecting the calculation of the microbial diversity. In contrast, the previous OTUs analysis utilized all the sequences in the samples. As shown in the study, a maximum of 157 and 635 Ananke TSCs were found with an average Ananke TSC comprising 0.6% and 0.2% of the data set in the human gut microbiome and in the freshwater lake microbiome, respectively. This result only represented a small subset of the microbial community. The clustering segregation is not apparent at all if TSCs are not manually highlighted in different colors. Ananke uses the non-phylogenetic metrics (STS) in the 3rd step to compute the distances between unique sequences. However, the phylogenetic metrics, such as UniFrac distance, were more common in the previous microbiome studies. More metrics should be available in the software (http://qiime.org/1.6.0/scripts/beta_diversity_metrics.html). The clustering parameter ε is hard to be chosen in 4th step of the software. The optimal number (or the robustness) of clusters can be estimated using the Calinski–Harabasz index and the silhouette analysis on k-means clustering. Ananke, as an unsupervised clustering though, needs to provide the measure to help the users determine the optimal number in a similar way. Figure 1, Figure 3, Figure 4, Figure 5, Figure 6 and Figure S3 have no units and no exact scales labeled in all the y-axes and in some of x-axes. The results of the study cannot be evaluated. The authors didn’t pairwise compare the results from Ananke to the ones identified in the original papers. The paired-end Illumina sequencing has been widely used, however, Ananke cannot analyze this type of the data.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ANANKE: TEMPORAL CLUSTERING REVEALS ECOLOGICAL DYNAMICS OF MICROBIAL COMMUNITIES Review round: 1 Reviewer: 3

Basic reporting: no comment. Experimental design: no comment. Validity of the findings: 1) The finding that 16S identity does not necessarily imply similar organismal ecology is not new, as the 16S rRNA gene has been shown since the discovery of its usefulness as a universal marker not to necessarily reflect the genomic diversity of the organisms (Abstract, Introduction). Yet, methods, such as Ananke, that aim at discovering finer ecological patterns are welcome. 2) The aim of the method is not clear: reading the Abstract and Introduction sections, it seems as if the authors propose to replace the current definition of OTU based on sequence similarity by clustering based on temporal profiles. Yet, the authors do not follow up in this direction in the Conclusion section (line 319), but seem to suggest otherwise its usefulness also for taxonomic purposes (e.g. lines 294-297). 3) The idea of OTU originates from the inability in microbiology to define the appropriate unit of evolution and selection. Therefore an "operational" definition is often used. Therefore one main question this study should discuss is whether a cluster of sequences defined based on a similar temporal pattern can constitute a unit of selection or evolution. How long should the time-series be then? How to standardize and compare across studies to compare those new temporally defined units? If we take several plant species for instance that are fluctuating similarly over time does this mean that they are under the same environmental triggers, under the same evolutionary pressure, etc.? I don’t think so, as this may be just a "guilt of association" case, especially if the time-series is short. Thus the important point of implications for evolution or selection should be discussed in the study. Also, the claim that “Ananke facilitates identification of potential microbial interactions" line 298 seems to go beyond what the tool can realistically provide (i.e. it can identify similar statistical patterns, not necessarily biological interactions). As a hypothesis generating/exploratory tool it seems however to be well suited. Time was chosen as the main factor to cluster sequences, but one could envision the same being done using an environmental gradient instead. This approach is also not new and is what constitutes the core of the ecological modeling and community ecology disciplines. 4) Methodological aspects. The method computes short-term changes in slopes between sequence abundances at successive time points. Looking at the methodological procedure several points need to be particularly checked: a. The sequence "abundances" are relative as they were standardized at each time point by sequence depth. Therefore, these abundances are not independent from each other anymore. The authors should consider including appropriate methods to deal with compositional data in their approach. b. The equation line 97 seems not to be scaled, i.e. that if longer time-series are used the distances will de facto increase. In most computation of distances, there is a scaling factor to make sure that the distance is scaled to [0, 1] for instance to allow meaningful comparisons. c. Lines 161-166, 214-245, 248-249: the procedure seems to arbitrarily exclude rare events or other types of sequences before computing further quantities. Therefore varying amount of singletons for instance which are then removed will lead to changes in relative abundances in the sequences that are kept for downstream calculations. Sequence depth is then affected. d. The authors based their work on the Möller-Levet et al. 2003 paper that clearly states that the method they developed is meant to be used for short time-series with uneven sampling points. Yet, the authors of Ananke used long-term (191 time points, lines 212-213) and rather evenly sampled data (daily samples; line 134). Is the approach then appropriately applied? e. In fact the authors used an OTU definition of 100% identity here because they investigated the changes in abundance profiles of unique sequences over time. What is the impact of organisms having several copies of the marker genes with potential sequence differences among the copies? (In the case of 16S rRNA operons, one would expect on average more than one copy per genome). f. There is a danger in proposing to "explore individual patterns" (e.g. lines 325-326), when no overall test for temporal changes when all data is considered altogether is performed beforehand. This would be akin to performing multiple pairwise t-tests instead of initially determining whether the ANOVA for a whole dataset is significant. Using an appropriate omnibus test is the only way to reduce the chance to observe false positives just by chance. There are some techniques to deal with time series for whole communities (i.e. multivariate time-series) and the authors should probably compare their approach to those established methods. Minor comments -Line 120: "Normal distribution" should be better explained because it is not clear whether the authors want to model a constant mean and constant variance over time (stationarity) or changes in variance over time even though the mean is constant (non-stationarity). In figure 1, other simulated "patterns" describe changes over time (x axis), whereas “normal” seems to refer to what happens on the y axis. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ANANKE: TEMPORAL CLUSTERING REVEALS ECOLOGICAL DYNAMICS OF MICROBIAL COMMUNITIES Review round: 2 Reviewer: 1

Basic reporting: no comment Experimental design: no comment Validity of the findings: no comment Additional comments: I appreciate the reviewers professional, complete, and satisfactory response to my initial review. The paper is well written, and Ananke is an excellent contribution. I commend them on the visualization code accompanying the manuscript that enables the user to easily plot and consider appropriate epsilon values for clustering.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: ANANKE: TEMPORAL CLUSTERING REVEALS ECOLOGICAL DYNAMICS OF MICROBIAL COMMUNITIES Review round: 2 Reviewer: 2

Basic reporting: no comment Experimental design: no comment Validity of the findings: no comment Additional comments: I do not have further comments and agree with the changes made by the authors.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: GENE EXPRESSION OF BENTHIC AMPHIPODS (GENUS: DIPOREIA) IN RELATION TO A CIRCULAR SSDNA VIRUS ACROSS TWO LAURENTIAN GREAT LAKES Review round: 1 Reviewer: 1

Basic reporting: no comment Experimental design: This study reports the transcriptome analysis of environmentally sampled Diporeia exhibiting differences in CRESS-DNA viral load. This report appears to be the first to explore transcriptional responses to infection with this virus in invertebrates. An analysis of differential expression was performed using RNA-Seq and standard methods, supported by qPCR data across 3 selected genes. DEGs between animals with high and low viral load are discussed in the context of their KOG classification and association with immune response. Interestingly, transcriptional changes in genes such as those involved in innate immunity were observed among the 3 sampling sites investigated. This is an interesting article within the scope of this journal that is well written and clearly presented. The authors provide a good amount of data to support their findings and I have been able to confirm the raw sequence submissions at NCBI. The main considerations relate to information surrounding the analysis of the RNA-Seq data in this study. Some detailed comments are listed below. Results and discussion It appears that there are multiple extant Diporeia species, can species be differentiated using COI gene analysis? It is unclear in the manuscript if the Diporeia sampled in the study are of the same species, though this was assumed with the mention of sub-species in line 189. The authors mention in Line 205 “the lack of clustering between samples is likely as a result of high variability between organisms”, this might be expected of the comparisons if the samples are distantly related. I feel that the term “haplotype clades” needs to be clarified in the context of defining relatedness between samples. Line 200: The reported contig N50 is rather low (290 nt), and read mapping results are quite poor (as low as 16% and not greater than 62%) as outlined in Table S4. It is possible that the reference assembly is still rather fragmented. The authors should clarify how this has been handled in the analysis (assembly metrics file seems to be missing – see General comments). Sample MI40(130) has a very low sequence yield compared to the other samples which may have introduced some bias to the DE analysis, unduly causing some dissimilarity with genes expressed in low levels. The authors should provide some support for retaining this sample in the study. Line 108: The authors noted that reads mapping to the SILVA database were excluded from the assembly. Were there reads mapping in high abundance to database entries for potential pathogens of Diporeia, that may suggest a possible co-infection? If so, this may be interesting to mention in the manuscript. It appears from Table S4 that a large proportion of reads map to the rRNA database, which seems to impact heavily on the overall sequencing depth. I would suggest that the authors offer a comparative example with another similar study in terms of their sequence yield. Validity of the findings: no comment Additional comments: Files and Figures Summary of transcriptome assembly and annotation statistics doesn’t appear to be included in the Supplementary Table S4 file as the file name suggests. This file seems to contain only information on raw read and trimming/mapping results. These results should be included. Supplementary Fig 2 appears to be missing? Can’t seem to locate the transcriptome stats mentioned in the manuscript. Supplementary file Raw_Ct_peerJ.d0.xlsx it appears should be named Supplementary Table 3. Minor grammatical correction required with the use of hyphenation being inconsistent for “over- and under-expressed”.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: GENE EXPRESSION OF BENTHIC AMPHIPODS (GENUS: DIPOREIA) IN RELATION TO A CIRCULAR SSDNA VIRUS ACROSS TWO LAURENTIAN GREAT LAKES Review round: 1 Reviewer: 2

Basic reporting: The MS is well written and easy to follow. The figures and tables are well laid out and i don not see any issues with this. Experimental design: The experimental design is good. Validity of the findings: Data is is statistically sound based on the methods used by the authors. Additional comments: I enjoying reading this MS. It is well laid out and relatively easy to follow. I like the concept of the MS and the aim to address the poor knowledge of the viral dynamics of LM29173 in Diporeia sp. I have a few minor suggestions and questions 1) Abstract and else where: 0-3x10E6. 0x10E6 = 0. So perhaps change to "up to 3x10E6" 2) Lines 135-139: can this be put in a supplementary table to make it easy to follow? 3) Line 330-331: change "(PCV1; Allan & Ellis 2000); TTV; Okamoto 2009)" to "(PCV1; Allan & Ellis 2000; TTV; Okamoto 2009)" 4) In your transcriptome analsysis did you find other CRESS DNA viral transcripts or RNA viral transcripts that may be co-infecting Diporeia sp and thus some of the up or down regulation may be as a result of a complex of viruses rather an only LM29173?

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: FILTERING OUT PARASITES: SAND CRABS (LEPIDOPA BENEDICTI) ARE INFECTED BY MORE PARASITES THAN SYMPATRIC MOLE CRABS (EMERITA BENEDICTI) Review round: 1 Reviewer: 1

Basic reporting: In the first line of the Introduction (L12) – the phrasing is rather awkward and should be changed to read “Parasites can be generalists that infect many different host species, or specialists that only infect a small number or even a single host species.” Additionally, the author should also provide a general reference for this statement, which can be one of the following: Poulin, R. (2006). Evolutionary ecology of parasites. 2nd ed. Princeton University Press. Schmid Hempel, P., & Schmid-Hempel, P. (2011). Evolutionary parasitology: the integrated study of infections, immunology, ecology, and genetics. Oxford University Press Loker, E., & Hofkin, B. (2015). Parasitology: a conceptual approach. Garland Science. At L25 poor reasoning was given for why that the unidentified nematode species might be a generalist. There are many non-host manipulating parasites which are specialists that infect only a few or even a single species. This should be modified to provide a better explanation for why the nematode might be a generalist. Experimental design: No comment Validity of the findings: In the Result section at L100, it would be very helpful for the read to also report the range (minimum – maximum) of parasite intensity. While this could be worked out by looking at Figure 2, this is not as exact as if actual numbers are provided. In the Discussion section at L146, the author stated “E. benedicti in this population is usual not only because…” – did they actually meant “unusual”? I have a suggestion about regarding the different patterns of cestode and nematode infection in the two crustacean species. Could there be some differences in the behaviour or other aspects of the two species’ ecology which makes one more susceptible or exposed to acquiring the infective stages of the parasites? This is especially worth considering since the unidentified parasitic nematode was absent in E. benedicti, which also infected far less frequently by Polycephalus larvae and in much lower numbers. For example, see: Koehler, A. V., Gonchar, A. G., & Poulin, R. (2011). Genetic and environmental determinants of host use in the trematode Maritrema novaezealandensis (Microphallidae). Parasitology, 138(1), 100-106. Differences in the body size of the two species may also be a contributing factor to why one species is more heavily (and in the case of the larval nematodes, just one of the species was infected) than the other. For example, see: Leung, T. L., & Poulin, R. (2008). Size-dependent pattern of metacercariae accumulation in Macomona liliana: the threshold for infection in a dead-end host. Parasitology research, 104(1), 177-180. Additional comments: The findings of this study are interesting and scientifically robust. The analyses used were appropriate and the interpretation of the findings was generally sound, therefore, once the author has made the suggested changes, it should be accepted for publications in PeerJ.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: FILTERING OUT PARASITES: SAND CRABS (LEPIDOPA BENEDICTI) ARE INFECTED BY MORE PARASITES THAN SYMPATRIC MOLE CRABS (EMERITA BENEDICTI) Review round: 1 Reviewer: 2

Basic reporting: This is a nice little study consisting of two parts. Firstly, parasite abundance was compared between two co-occurring sand crab species. Secondly, behavioral tests were conducted on one of them to investigate whether the cestode Polypocephalus sp. induces behavioral changes, which may indicate parasite induced trophic transmission to the final host. This is of particular interest because the parasite has been linked to behavioral changes in a different crustacean (Carreon et al. 2011), and the behavioral tests on Lepidopa benedicti and the quantification of parasites in that species provide the main story of this paper. There are some issues related to the comparison of parasite data between the two species, and there are some potential issues related to the interpretations of the behavioral data, which I will address in more detail in the relevant sections of my review below. 1) The study system is well introduced. However, the concept of PITT deserves to be defined and introduced earlier in the Introduction followed by a general reference. In this case, Lafferty and Shaw, 2013 would be a suitable reference. Moore, 2002 is another good general reference (book). 2) Line 22-23: I think it is reasonable to assume that species sharing the same habitat are exposed to the same set of parasites. However, there could be many physiological and behavioral reasons why some parasites successfully infect one species and not another species. Therefore, it is a stretch to assume that co-occurring species harbor the same set of parasites. I recommend that you rephrase the sentence accordingly. I think it is important to not produce a too simplistic view of host specificity. Many other factors than the ability to change host behavior could be important. For example, parasites, which have to negotiate a potentially strong immune response are often host specific. 3) Be clear and consistent in the use of terminology. Parasite intensity refers to the number of parasites in infected hosts only. If uninfected individuals are included (which I believe you do), then use the term parasite abundance, which includes all individuals examined (including uninfected individuals). Refer to Bush et al. 1997 for a more detailed outline of different parasitology terminology. 4) L. 108-110: This sentence belongs in the Discussion (not part of this study). Figures 5) In Figure 3 I believe that the scale bar is incorrect in either A, or in B and C (the size of the parasite is very different between photos). Also in Fig. 3, I would only provide two photos, one from Lepidopa benedicti and one from Emerita benedicti, each of them indicating parasites in tissue by arrows. 6) Fig. 5. In the Figure legend I suggest that you define the abbreviated term SEG. Sub-esophageal ganglion may not be known to a wider audience. 7) There are quite a few typos/spelling mistakes throughout the manuscript, which need to be addressed (e.g. lines 41, 57, 80, 114, 141, 146, 150, 160, 168, 169). Experimental design: 1) The research question states that behaviour would be tested in both species (L. 50-53). However, this was only done for Lepidopa benedicti. Rephrase the statement accordingly. 2) It is problematic that you compare nematode data collected at different times between the two sand crab species. This is simply not meaningful in many cases since parasite abundance may vary greatly both spatially (within even a few meters) and temporally. My recommendation is that you remove data on the unidentified nematode (Fig. 2). You can mention in the Results section that you did not find the nematode in Emerita in this study, and then refer to the previously published study reporting nematode infections in Lepidopa (Joseph and Faulkes 2014) as long as you do not directly compare the two findings. 3) The methods are not explained in sufficient detail for replication. For example, how and where specifically were the animals collected? (it only says that they were collected). Also, how were Emerita dissected to examine for the presence of nematodes? 4) L. 70-72: A minor comment, but behavior alterations may not necessarily be intensity-dependent. Therefore, presence/absence of a parasite may still provide an indication of if a parasite may change host behaviour. Validity of the findings: 1) My main concern regarding the interpretation of your results is that the effect of crab size on behavior was not investigated. You report that crab size was significantly and positively related to the number of Polypocephalus per individual (Fig. 4). That could also mean that the negative relationship between the number of Polypocephalus and digging time observed (Fig. 6+7) was influenced by the effect of crab size. Maybe you can use crab size as a co-variate in your analyses to correct for the potential confounding effect of size. 2) You use parametric tests to test the relationship btw size and parasite numbers, and digging time and parasite numbers. To me it looks like your data violate the assumptions of homogeneity of variance in parasite numbers (y values) across x values (size). It is clear that variance in parasite numbers increase with increasing size of the sand crabs. I therefore suggest using a Spearman rank test instead of a Pearson correlation or Linear regression analysis. You may also be able to transform your data so that you can still use a parametric test (especially if you want to include size as a covariate). 3) I wonder if you measured the dimensions of the larval Polypocephalus in any of the two host species or in white shrimp? Molecular techniques are necessary to determine the identity of Polypocephalus collected from different hosts, but morphological measurements may indicate obvious differences. 4) Along the lines of the point above, I think you should mention that white shrimp generally occupy a different habitat from sand crabs (sea grass beds as juveniles and deeper waters as adults in contrast to the swash zone of sandy beaches), which may suggest that we are dealing with a different species of Polypocephalus. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: FILTERING OUT PARASITES: SAND CRABS (LEPIDOPA BENEDICTI) ARE INFECTED BY MORE PARASITES THAN SYMPATRIC MOLE CRABS (EMERITA BENEDICTI) Review round: 2 Reviewer: 1

Basic reporting: I am satisfied with the revised version of the manuscript and recommend that it be accepted for publication. Experimental design: no comment Validity of the findings: no comment Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: FILTERING OUT PARASITES: SAND CRABS (LEPIDOPA BENEDICTI) ARE INFECTED BY MORE PARASITES THAN SYMPATRIC MOLE CRABS (EMERITA BENEDICTI) Review round: 2 Reviewer: 2

Basic reporting: I have no further comments Experimental design: No further comments Validity of the findings: No further comments Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE DARK CUBE: DARK CHARACTER PROFILES AND OCEAN Review round: 1 Reviewer: 1

Basic reporting: The paper is a standard descriptive study with all appropriate details included. Experimental design: While not experimental, the research design is clear. I am troubled with the assumptions made in the cube design. While the authors notes the limitations of median splits in brief I think more needs to be said about problems with assuming traits are categorical as opposed to continuous and how this approach is better than a latent variable approach. The authors repeatedly allude to the unification or dark dou but it is unclear whether the present analysis is best here. The important point that might be missed in relation to the former is that it is the SHARED variance not the unshared variance that is of interest to many DT researchers. In reference to the second point, the two factor solutions have reliably been detected in very large samples which introduce more correlated errors--errors that M and P people may make systematically more than N people. Validity of the findings: The findings are purely descriptive in nature. This study does not advance any DT theory is not one. The use of the B5 is a descriptive tool that allows us to understand the DT traits. As the B5 have become the language of all personality psychology, framing the DT in terms of the B5 helps others understand what is meant by the traits. However, all we really have here is correlated semantic agreement. The study does not advance any theoretical model and instead adds to the noise of B5 studies on the DT. While the authors have adopted a novel method, the results should not be oversold as resolving any conflict or question. Additional comments: I find it notable that you write an entire paper on the DT and not cite me one time. This is especially problematic as I was the one who first postulated the presence of a DT factor; a finding that has been replicated both by my work but also other work by Figueredo et al. (2015; Evolutionary Psychology). The primary limitation of this paper is that is lacks any real thoeretical heft. It is completely descriptive in nature but the authors seem to overstep their mark here.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE DARK CUBE: DARK CHARACTER PROFILES AND OCEAN Review round: 1 Reviewer: 2

Basic reporting: The manuscript ’The Dark Cube: Dark Character Profiles and OCEAN’ by D. Garcia and F.R.G. Moraga addresses an important issue in the research of the Dark Triad (DT). Namely they argue to investigate the possible combinations of high and low DT traits in relation to the Big Five. This line of research is of high importance because the unity of DT traits has been questioned by several authors. The current study tries to make good for the lack of consistent relationships between DT and the Big Five throughout the literature. The introduction gives a concise but clear outline of the problem and of previous research in the field. However, I found some of the phrasing somewhat misleading throughout the manuscript. I will list all three of them here. (1) Labelling the mnp profile (low Mach, low narcissism, low psychopathy) as ‘agreeable’ might be confusing, because it posits agreeableness as an a priori anti-thesis of the DT (which is in fact the case in many studies). I would suggest to label mnp as ‘beneficent’ in contrast to the label of MNP (i.e. maleficent). (2) Authors highlight the importance of investigating non-linear relationships between DT and OCEAN. I found it also misleading, because linear relationships are investigated (comparing two groups can only be linear). At the same time, I agree with authors about the importance of investigating conditional or moderated relationships. I would suggest two use on of the previously mentioned terms instead of non-linear. (3) Although it has been already published (Garcia and Rosenberg, 2016) – and I can accept if authors chose to stick with their original phrasing –, labelling the character cube by Cloninger as ‘light’ might be also confusing. Cloninger’s TCI measure is frequently used in the assessment of personality disorders that are mainly based on character traits (especially SD and C). Simply omitting ‘light’ would easily solve this problem. Moreover, please spell-check the manuscript throughout. E.g., p. 11 row 3 would go correctly ‘… is negatively associated to …’. Experimental design: The study used a self-report cross-sectional design. The limitations resulting from this design should be mentioned somewhere in the Limitations and concluding remarks section. Authors should also report, which statistical software they used for the analyses. Validity of the findings: The findings in their current form cannot be considered as valid. I will list my concerns below in detail. (1) Dividing the sample into groups representing different Dark Cube profiles leaves the authors with two problematic groups. MNp consists of 11 participants and mnP consists of 17 participants. Using these groups in any comparison makes the validity of the results highly questionable. Suggested solution: authors should at least add a dimensional test of moderated moderation (or three-way interaction) for DT traits. This would be the same statistical approach as in Garcia and Rosenberg (2016). Using modules like e.g., PROCESS for SPSS would also enable authors to trace the conditional effects of DT traits on OCEAN. (2) Since gender distribution is uneven in the sample and both DT and OCEAN are gender-sensitive, authors should at least report the gender distribution for each profile group. They might be also willing to control for effects of gender (and perhaps age) in statistical analyses. (3) Although it might not fit in one table, but somehow group descriptives (M and SD) should be reported, maybe along with Cohen’s ds as indices of power. Additional comments: I would be really happy to see your manuscript published, because it could stir up the water in DT research. However, half of the sample categorized either as mnp or as MNP leaves me unpersuaded whether it is correct to talk about a Dark Cube. This distribution shows that M, N and P are highly correlated, whereas a cube has three orthogonal dimensions. Maybe establishing national standards for SD3 on a representative sample would help with categorizing anyone over 50 T as high on a trait and equal to or below 50 as low on a trait. It would give us a realistic picture about the proportion of any profile in the population (I would be really sad if – as your sample shows – we would have 25 % of the population as maleficent).

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE DARK CUBE: DARK CHARACTER PROFILES AND OCEAN Review round: 1 Reviewer: 3

Basic reporting: For clarity purposes the grammar and spelling should be checked. It would have been also really helpful, if the data and the code were attached. Without those it makes it hard to comprehend the outcomes of the paper. The introduction makes really clear that personality can be understood as a complex interplay of traits. But the authors fail to make really clear what the advantages of the dark cube actually are, and how introducing personality types can account for that complexity. While checking some citations I came across some problems also. A paper by Lee & Ashton (2013) is used to give an example where a certain correlation pattern between the Dark Triad and the Big Five can actually be found. This article doesn’t even contain measures of the Dark Triad, but it is an article about validity of the HEXACO and NEO-FFI, so it is not relevant to the topic. And just for example, another paper used to underscore the inconsistency of correlation patterns between the Dark Triad and the Big Five by Jakobwitz & Egan (2006), used different measures of those personality traits than the paper it was compared to (Paulhus & Williams, 2002). What I mean with that: differences in the correlation patterns can stem from a lot of different sources. Not mentioning or discussing those different sources seems unscientific. The introduction doesn’t go enough into detail about the concept and the advantages of the Dark Cube, and needs to reassess the literature itself, as well as the way those publications are used. The structure of the article does not conform to an acceptable format. Results and discussion should be handled separately. It is very important to first just show the results without any interpretation, and then interpret those. Experimental design: A big part of my concerns with this paper is about the methods, and statistical analyses used. There is no explanation given in the method or result section why certain analyses where used. I would have liked to see for example an explanation given to why the median but not the mean was used to split the groups (as well as for other analyses: why a t-test?). This would support clarity and transparency. The reliability measures given for the Short Dark Triad (SD3) are not sufficient, please provide an internal consistency score for the whole measure also. It is not made clear, if the participants were speaking Spanish or not. Given the Spanish BFI measure I would think so, but there is no Spanish SD3 measure cited or mentioned. Please provide information about this issue. I understood this paper as a way of showing researchers who work with the dark traits a new and different way of understanding and analysing the Dark Triad. For that it would have been important that the authors also show the actual correlations they found between the Dark Triad and Big Five, and compare them to the way they analysed. Not showing those results is highly problematic. Some other information about statistical analyses that I missed were how the authors accounted for different sample sizes for the t test. Also there were no effect sizes provided, which are just as important as significance levels. The explanation given in the discussion why it is statistically sound to divide 8 different personality types by the median of a normally distributed trait is not sufficient at all. The discussion part doesn’t really focus on trying to find explanations why the results are the way they are. One really interesting result (that Agreeableness seems to be less connected to the dark traits than expected) was just mentioned and discussed at all, even though it is a highly controversial point to be made. Validity of the findings: Validity of findings is already discussed in first two parts of review. Additional comments: No additional comments

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: THE DARK CUBE: DARK CHARACTER PROFILES AND OCEAN Review round: 2 Reviewer: 1

Basic reporting: The revised manuscript uses professional English. Introduction and discussion are provided with sufficient references and context. Article structure is according to standards, figure and tables are ok. The submission is self-contained. Experimental design: Original research fitting into the Aims and Scope of PeerJ. Research question is well defined, addresses an important gap in Dark Triad Research. Investigation is in line with APA ethical standards. Method is described in sufficient detail for replication. Validity of the findings: Data are robust and statistically sound. Conclusions are well stated. Speculations are clearly identified as such. Additional comments: I have all my comments answered in sufficient form.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: SPONTANEOUS REOCCURRENCE OF “SCOOPING”, A WILD TOOL-USE BEHAVIOUR, IN NAÏVE CHIMPANZEES Review round: 1 Reviewer: 1

Basic reporting: It is clearly written and argued, and I have no further suggested improvements on the basic reporting. Experimental design: This is an original and important paper that provides a new theoretical basis for tool behavior in primates (and other animals as well, I would say). I can envision this study as a jumping off point for similar, intriguing research - as the authors note, more studies are needed to identify which behaviors are within great apes’ spontaneous capabilities. Validity of the findings: This is a sound study, and I mainly had minor questions. However, per Lines 109-115: This seems a post-hoc proposition based on the research presented here. I would feel more comfortable seeing this in the Discussion section of the manuscript. I don’t think it would take away from the authors’ arguments, and it follows their findings, which are convincingly argued. That said, however, “at least two independent target-behaviour-naïve individuals” does seem subjective given the results of this study. Again, I am persuaded by the authors’ arguments, but ideally would they say that the majority of the population would exhibit such behavior (i.e., if logistics such as those preventing the other subjects in the current study from exhibiting such behavior were not in place)? On a population level, this would seem to be a better argument, especially as the authors state in the Abstract that individual-level behaviors can help achieve population-level patterns. Perhaps it is beyond the scope of this paper to ascertain what proportion of a population would exhibit such individual behaviors, but it is something that is at least relative to their theory. Additional comments: Line 64 – as far as neglecting the origins of behavior, might this be more along the lines of a rarity in terms of observing novel behaviors? (rather than neglecting discussion of origins) That is my understanding of the records of novel tool use behaviors that appear and then spread (or not) among wild chimpanzees at least. Line 76 – could the behaviors be listed here with the relevant references? Line 77 – “…but are also spread” seems to contradict the following sentence – at least in the way I read it. I’m not sure if rewording would help clarify… Line 80 – “socially mediated serial reinventions” is a mouthful (but clear and accurate) - If this is used often, I would note “(SMSR)” here, following ‘socially mediated serial reinventions’ and use the acronym throughout Line 83 – Does ‘the cumulative nature of our culture’ necessarily include active teaching? If this is implied, I would qualify this phrasing or somehow further elaborate. If this statement is just a broad one, I’m not sure how helpful it is to the reader as it is written – but perhaps I’m not following the connection between our cumulative nature and specifics of the ratchet effect (which I am familiar with) Line 88- do humans also have ZLS? Line 92 – should “spontaneous teaching” be “spontaneous active teaching”? Line 93- could the authors references these findings again here (per “see above”) Line 100- perhaps “can then spread” Line 114 – could the authors clarify what they mean by enculturated? Line 120 – could the authors clarify what they mean by unexceptional? Lines 123-129 – per weaknesses of previous studies…I’m wondering if this might be better suited to the Discussion? Also, perhaps the authors should explicitly state how their study is providing baseline data – I’m not following their argument to the point at which I think they would like the reader to, given the pointed nature of this section… Line 158 – change “decided to divide” to “divided” Line 194 = spell out 14 at the beginning of the sentence (usually) Line 197 = add comma between ‘areas’ and ‘which’ Line 206 – I’m assuming this is because of zoo management…? Could that be used instead of ‘local restrictions’? Line 208 – can the sentence ending on this line be qualified with, “…during testing.”? Lines 219-220 – this is interesting in that a challenge will be to actually find naïve chimpanzees for most tool tasks!

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: SPONTANEOUS REOCCURRENCE OF “SCOOPING”, A WILD TOOL-USE BEHAVIOUR, IN NAÏVE CHIMPANZEES Review round: 1 Reviewer: 2

Basic reporting: English is correct. Literature covered is correct, but some refs are missing (see comments below). Structure is fine. Results: it's unclear whether the authors truly test what they are intending to test, as it is unclear whether chimps are truly 'naive' to the task of using a stick as a tool (see below). Experimental design: Aims and scope: yes Research question is well defined. Investigation is performed ethically. Methods are described with sufficient details to be replicated (see below for comments). The major problem with the claims of the article is that it is absolutely necessary that the chimps are completely naive to the task. This can be easily tested with presenting them with an apparatus containing a valuable resource which is out of reach. Some populations in Africa have been found to be completely naive to stick use. If the chimps already know that sticks are potential tools, there is not much of a big cognitive jump to use a stick for a different task. Validity of the findings: The novelty is not clear as the authors have already published similar claims before, albeit not on this particular behaviour. There is not much to comment on the data as they are relatively simple. Conclusions are speculative at this point because it is unclear whether the ZLS is tested by the current design (particularly because of the choice of the chimp subjects). I believe the discussion should be organized about chimp tool flexibility (which is demonstrated by this experiment). See below my comments. Additional comments: I had already the opportunity to review this article for a different journal, providing the authors with a detailed review and comments to be addressed. Sadly, the authors did not address much of what I wrote so I am re-enclosing my former review, with the hope that this time, the authors will actually pay attention to my comments. Considering that the other reviewer for that journal had similar points to mine, I would assume these must be addressed if the paper is to be considered for acceptance. The lines have been re-adapted for convenience, and comments on changes by the authors or novel information have been added between ticks. This article claims to bring a welcome test of the Zone of Latent Solutions (ZLS), applied to tool use. In general, I thought the article was quite well written, although the review of the literature appears too biased towards the hypothesis defended by the authors (that is, chimpanzee cultures result only from individual learning), rather than giving a balanced account of the current literature and debates and thus would benefit from being re-balanced. As such, this study follows on previous work by one of the authors, aiming to document whether ‘naïve’ apes can replicate behaviour observed in the wild, presented as ‘reinnovation’ or ‘reinventions’. First, I am glad to see the ZLS explained more than in the first study introducing the concept (Tennie et al. 2009). Particularly, I thought that highlighting two particular aspects of transmission, that is transmission of the form, and transmission of the ‘idea’, was quite convenient, especially to further the debate about the ZLS. However, I am unclear whether this particular study really tests the ZLS, rather than innovative abilities in chimpanzees. One major problem, I believe, is that the tested chimpanzees were already familiar with sticks (and bread). As such, their use of sticks to obtain floating bread is not surprising. The authors also make a whole case of why this result would be relevant to finding of ‘chimpanzee cultures’ in the wild. However, it is unclear how this study can have consequences on interpretations in the wild. Indeed this study really investigates flexibility in chimpanzees (how to make use of one’s knowledge to solve new problems), rather than the ZLS itself. In this respect, the abstract and main conclusion should be rewritten in this spirit, rather than focus on attacking the claims of chimpanzee culture in the wild, which the authors cannot address with this experiment. My comments follow the text: Abstract -L17-19: I find this statement quite confusing. First of all, down to the individual level, every behaviour learnt (including socially) depends on individual learning. Then, this learning may be influenced by social factors or not. As well characterized by Fragaszy & Visalberghi 2001, social learning is better defined as ‘socially-biased individual learning’ (see also Heyes, 2012, JCP for a more recent review). I think in the article, the authors make a better job of explaining what they mean. If I understood well, for them, ‘individual social learning in humans’ influences the form of the behaviour acquired. In contrast, in chimps (and other animals), at best, they get the idea of what they should do from others (thus socially) but have to re-invent it themselves. Why not explain it like that already here? The organisation of the abstract is also slightly awkward, with much of devoted to explaining what the ZLS is, rather than actually introducing the study. The abstract may benefit from being rewritten more along the lines of standard abstracts. Introduction -L54-55: “ecologically appropriate chimpanzees” does not seem appropriate to me, particularly according to the definition given by the authors. Only wild chimpanzees are technically ‘ecologically valid’ chimps. At the same time, one can make a case that captive chimps’ environment is their ‘natural habitat’. Whether captive environment can however translate into valid representation of wild habitat is much more debatable. I think the authors should characterize better what they mean by this expression, but also acknowledge that captivity is a poor substitute to the wild. -L56-60: I think this sentence should be rephrased, mentioning that some studies fail to find evidence of imitation, while others find evidence of imitation, leading to the conclusion that chimpanzees have a range of possible social learning mechanisms at hand, including imitation, but that they don’t rely on the latter all the time, and particularly not as much as humans. The way it is now is too biased toward the authors’ belief that there is no imitation in chimpanzees. -L69-72: This is not correct. According to the authors of the article mentioned, social learning did play a major role (85-99% see abstract) both in the ‘spread’ and ‘the appearance of the behaviour of others’. Thus the sentence has to be modified to reflect this. Additionally, while the behaviour may have been rediscovered by one individual, this is because this individual picked up a moss-sponge herself, and the authors conclude that the most likely explanation for this transmission was some form of stimulus enhancement. This is social learning nonetheless. However, I believe this could go along the lines of what the authors discuss later on (in terms of low-level social learning mechanisms). 'Finally, the Zuberbühler group has produced a novel study that came recently (Lamon et al. Science Advances 2017), which brings further evidence of social learning. I thus encourage the authors to review their description of this particular set of studies if they want to make a point about social learning in chimpanzees in the wild.' -L77-80: Although I believe this wording here (‘we argue….’) is clearer than in the abstract, does this statement not amount to the old Tomasellian argument about imitation vs the rest? Technically, only imitation would allow the transmission of the form of the behaviour (but again, it’s unclear whether imitation is really different than other social learning mechanisms, see Heyes 2015, Dev Science). In general, I would agree that it is interesting to discuss these two aspects of social learning, one that is more concerned with spreading the form of the behaviour, and the other one more concerned with spreading the idea, so to speak, but in that case, why would the authors not apply the same argument to humans? We could also say that the spread of the behaviour in humans results from socially mediated serial reinventions, all based on an incremental scale (based on each generation’s pool of knowledge). -L92-95: I do not understand why the authors jump straight from imitation to individual learning (rather than, say, emulation, if that is what apes do according to them). As per their summary above, they were hinting at great apes being emulators rather than imitators. This can be debated, but let’s agree to that for the sake of the argument. Then, it is more parsimonious to assume that emulation is the main motor that drives the emergence of tool-use behaviours in apes, rather than individual learning. Again, we are still faced with the same problems: first social learning is, at its core, based on individual learning, but socially-biased. Second, the emergence, at the group level, also depends on innovation abilities of single animals. The authors need to make a better job of explaining their stance on these two topics (I have felt that the innovation side has been quite occluded in the intro). 'Compared to the former version, the authors have added a statement between brackets about the fact that there is a debate about imitation. While this bracket may be useful at a different place in the manuscript, this is not the point of the main sentence to which the bracket is connected and which states that “individual learning”, rather than any sort of social learning is actually at work in chimpanzees, which is not true!' -L99-102: I’m unclear on how low social learning mechanisms may homogenize a given ‘form’ of behaviour so the authors need to develop this more. This is only possible if the pool of possibilities is itself quite limited, and thus cannot grant much diversity at the end. If not, one would expect to see more diversity in the use of one particular tool, equivalent to chance level to select a given option in the use of that tool. -L103-104: I think the authors need to somehow come clean here. At the beginning of the intro, we are led to believe that the authors favour individual learning for the form, then social learning (albeit low level) for the spread, which could somehow characterize the ZLS. But now, it seems it is only the ZLS is only about individual learning. Which side are we supposed to take? -L116-139: This paragraph should go higher in the introduction. Possibly before L105. In general, it could probably be beneficial to have some discussion about individual learning vs innovation too, and how they are taken into account in the ZLS framework (see my point above). Methods -L214-231: I believe the fact that the chimps already knew about stick use is actually a major drawback to the claim that the chimps are naïve in this study (see below). In fact, it would be indicated to summarize here or in the supplemental info how much knowledge the chimps had about stick use (e.g. termite mounds etc…), as per the carers’ statements. 'The authors still haven’t answered my comment about the kinds of ‘stick knowledge’ that was available to the chimpanzees. If the chimps were already used to use stick to recover food (e.g. in a termite mound or from a pot of some sort), there is no massive cognitive jump to use this knowledge to recover food from a body of water. I still would request a detailed description of the tool behaviour of the chimps prior to the experiment.' Results -L269-271: I am not clear with the statement about copying here. If we follow the argument of the authors in the introduction, then it’s all individual learning anyway, or simple stimulus enhancement, right? So why not analyze the behaviour of other individuals? This could possibly bring some data in favour of the authors if all individuals demonstrated a different technique to reach the bread. -L274-275: Even though they discuss it at the beginning of the Methods, I am not clear why the authors use the word ‘re-invent’ here. They should rather use the word ‘innovate’. The chimps did not re-invent scooping, if we assumed they were naïve before, so they innovated it. I believe the authors need to address the difference between ‘innovating’ and ‘reinventing’ in the Methods. If not, they should use the word ‘innovate’ throughout the manuscript. -L276-279: I am wondering about how much the fact that the chimps had to go through the mesh led them to be cautious and ‘gently rotate the wrist’. Basically, where their hand movements highly constrained? -L285-292: Considering this statement, I do not think the authors can claim above that the behaviour was indistinguishable from the wild counterpart. -L314-318: Usually success time (or any duration) is given +/- Standard Deviation. Discussion -L338-339: I don’t think any behaviour require ‘social learning’ to be ‘invented’ or ‘reinvented’. It requires innovation abilities by definition! So the authors are fighting a strawman here. The sentence should be modified. -L341-343: I think one major caveat here is that the chimps did know about stick use before, and most especially, stick use to obtain food. It has been shown that wild chimpanzees have much difficulty to innovate stick use to obtain food when the behaviour is not present in their behavioural repertoire (see work by Gruber et al. 2011). Thus it is unclear whether the chimps in the present study needed in fact any learning at all (social or individual). They knew about sticks, possibly knew how to use sticks to get food (this absolutely needs to be cleared in a reviewed version of the article), they knew about bread, and thus, what they innovated at best was to use a stick to recover a food source. For the record, I am not saying that the wild chimps of Bossou did anything differently. They probably already knew stick use, and may have tasted the algae without using sticks at first. But this particular point needs to be addressed. It is fundamental to the argument of the authors. -L352-356: Here there is a third definition/characteristic of ZLS: that all innovations are latent solutions. But then the whole ZLS argument should not be about social learning, but about innovation itself. But then the intro must be reshuffled to introduce discussion about innovation. In a nutshell, we are presented with several definitions/exemplifications of the ZLS. The authors need to pick one and stick to it! -L372-375: I am unclear about why, evolutionarily, apes (I assume the authors exclude humans here?) would evolve enhanced individual learning, particularly considering that some populations of chimps and orangutans, and most populations of bonobos and gorillas are fairly limited in tool use. Also, why would have this disappeared in the human lineage in favour of social learning? In fact, humans are probably the best individual innovators, which is one founding block of cumulative culture (although children are, counter-intuitively, quite bad at it..). -L379-382: The authors are still misinterpreting the chapter by Humle et al. The point of these authors relates to the similarity between individual techniques relating to proximity between specific individuals. It is not about “individual differences in single actions” but rather individual similarities. Therefore, the authors have to replace the word “differences” by the word “similarities”. -L419: Again, it is unclear to me whether non-stick-using chimpanzees would actually be able to innovate this behaviour. This has been tested in the wild (see work by Gruber et al. 2009, 2011) and chimpanzees failed to “reinvent” the behaviour. I believe the authors should give reasons why they would think there was any innovation at all at play in their study. If anything (and taking into account the ‘slide’ behaviour) the chimps demonstrated flexibility in their behaviour and use of their old knowledge to solve a novel problem. -L428-429: This is actually quite a ‘big’ caveat, because it seems that the ZLS would suggest ‘blind’ acquisition of the behaviour. That is, as long as some kind of low-level social learning mechanisms appear, the animal would not have controlled on whether to decide to do the behaviour or not. In a nutshell, it would necessary have to display the behaviour, which would then automatically ‘spread’. -L432: I am quite uncomfortable with the use of ‘re’ something everywhere in the paper. While the authors mention it in the methods, they do not justify it. What is actually the case for it? ‘reinnovate’ to me, would amount to chimps having had the behaviour in a given population, which disappeared, and which subsequently re-appeared. This would qualify as ‘re-innovation’. But that zoo chimps innovate the same behaviour as displayed in the wild does not qualify as ‘re-innovation’. 'I’m still unclear about the use of re-invent, re-innovate, etc… Can’t the authors provide clear definitions for each of the words they use here?L189-192 are uninformative regarding this.' -L441+: While I congratulate the authors for addressing potential criticisms, I feel like the most important one that has not been addressed is the fact that chimps all knew how to use sticks, irrelevantly of whether they saw or not the actual scooping behaviour. That is, they already knew that sticks are tools that can be used to recover food. Then there is not much needed to ‘rediscover’ the behaviour, under the form of using a particular motion (driven by the necessity to drive the stick through the mesh) to recover… food. The authors need to make a case to explain why they believe they actually witness any kind of ‘reinvention’. 'The paragraph added L449-454 does not address the fact that chimps may have some knowledge about the use of sticks to obtain food, for instance retrieving honey or jam from a termite mount, or from plastic tubes filled with food, which are quite often provided by zoos for enrichment. The authors have to state whether chimps had access to this kind of enrichment (through a list of stick behaviour they might be familiar with), and if that’s the case, how they can address this caveat. ' -L494-496: Once again, I am not convinced that the authors address the question of high fidelity social learning with their experiment, rather than the innovation abilities of the chimpanzees. I believe this distinction will have to be addressed in the reviewed version of this article. Final comment on the discussion: I believe it might benefit from having subtitles or sections. At the moment, the reader may become a bit lost when reading the whole thing at once. Thus I would suggest re-organizing it logically and/or introduce section titles.

You are one of the reviewers, your task is to write a review for the article. You will be given the title of the article, the number of the round in which the article is located, and your order among the reviewers.

Title: SPONTANEOUS REOCCURRENCE OF “SCOOPING”, A WILD TOOL-USE BEHAVIOUR, IN NAÏVE CHIMPANZEES Review round: 2 Reviewer: 1

Basic reporting: See below Experimental design: See below Validity of the findings: See below Additional comments: I thank the authors for addressing (in details!) my comments. First, the article reads much better and it is much easier to follow the line of thoughts of the authors. While I still fundamentally disagree with the notion that chimpanzees were 'naive', as they had knowledge of sticks, I appreciate that the authors made an effort to pin down exactly what they are after, that is testing whether the form of a behaviour can be reinnovated individually. I would probably request a few more things added in the intro and discussion (this, or I guess the reviewer-author exchange might be published together with the article so that it outlines the various opinions regarding the data...). 1) Regarding the intro, and looking at the comments, I still believe there is some disagreement about what is social learning and what is individual learning. When I wrote that even in humans,'down to the individual level, every behaviour learnt (including socially) depends on individual learning', my point was to stress out that each individual, to develop the behaviour, must learn it individually. In effect, an individual must have the physical and mental capacities to develop a given a behaviour to be able to develop them. I agree with the fact that individuals may not have the 'innovative abilities' to invent a new behaviour anew, but it does not make sense to write that 'many learnt behaviours are outside of the individual-LEARNING capabilities of each individual'. This would basically mean that you need to be several individuals to be able to acquire one particular behaviour (maybe this is the case for bird flocks, but not for humans, as far as I understand it). Rather, what the authors mean is that the behaviour we display now in 2017 would never have been innovated by someone who say lived in England in the XVIIIth century. But down to the individual, it's still their individual learning capacities that allow her/him to 'learn' a behaviour. To take the tango example given by the authors, while the moves of tango result from cumulative culture (so learning the moves needs social learning), the individual has to use individual-based learning mechanisms such as working memory to register and display the moves in the right order. So again, social learning is then socially-biased individual learning. See Heyes' arguments. 2) Second, regarding the discussion, I believe it would be worth discussing in the section about "not naive chimps?" the splitters vs lumpers argument, which the authors use to their advantage here. This gets back to my point that the authors demonstrate little more than chimps being good at transferring knowledge from one task to the other other, which contrasts with findings from the field. If one analyses each behaviour as a different innovation, then, clearly there will be many behaviours falling into the ZLS if an animal already knows a closely related behaviour (e.g. stick use for algae scooping). But that does not make a strong argument for ZLS as an hypothesis. In effect, that makes it weaker. In contrast, if an animal develops stone cracking from scratch (and without any human help), then it makes a much bigger case for the ZLS. A few more minor comments: Introduction: -L129: A dot is missing. Discussion: -L419: I am not sure you should cite the Koops et al. 2014 paper here, if your point is to say that both opportunity and necessity interact to lead to the appearance of a tool behaviour. In contrast, Koops et al. 2014 state that only opportunity, but not necessity plays a role in the appearance of tool use. Rather, you should either cite the original book chapter (Fox et al. 1999, Intelligent tool use in wild Sumatran orangutans) or more recent papers that have advocated for such a view combining the two hypotheses (e.g. in capuchins: Moura & Lee, 2015 book chapter Necessity, unpredictability and opportunity: an exploration of ecological and social drivers of behavioural innovation; in chimps: Gruber et al. 2016 eLife). -L441-442: I think the authors could mention here that innovation skills are in fact quite limited in young human children compared to their well demonstrated social learning skills (e.g. Nielsen 2013).