{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from IPython.display import display_html\n",
    "\n",
    "import logging\n",
    "import warnings\n",
    "import re\n",
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pickle\n",
    "import pickle\n",
    "import requests\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from rdkit import Chem\n",
    "from rdkit.Chem import AllChem\n",
    "from typing import Literal, Union, List, Dict, Any, Callable\n",
    "from collections import defaultdict\n",
    "from tqdm.auto import tqdm\n",
    "from rdkit import RDLogger\n",
    "\n",
    "RDLogger.DisableLog('rdApp.*')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def set_global_logging_level(level=logging.ERROR, prefices=[\"\"]):\n",
    "    \"\"\"\n",
    "    Override logging levels of different modules based on their name as a prefix.\n",
    "    It needs to be invoked after the modules have been loaded so that their loggers have been initialized.\n",
    "\n",
    "    Args:\n",
    "        - level: desired level. e.g. logging.INFO. Optional. Default is logging.ERROR\n",
    "        - prefices: list of one or more str prefices to match (e.g. [\"transformers\", \"torch\"]). Optional.\n",
    "          Default is `[\"\"]` to match all active loggers.\n",
    "          The match is a case-sensitive `module_name.startswith(prefix)`\n",
    "    \"\"\"\n",
    "    prefix_re = re.compile(fr'^(?:{ \"|\".join(prefices) })')\n",
    "    for name in logging.root.manager.loggerDict:\n",
    "        if re.match(prefix_re, name):\n",
    "            logging.getLogger(name).setLevel(level)\n",
    "\n",
    "\n",
    "# Filter out annoying Pytorch Lightning printouts\n",
    "warnings.filterwarnings('ignore')\n",
    "warnings.filterwarnings(\n",
    "    'ignore', '.*Covariance of the parameters could not be estimated.*')\n",
    "warnings.filterwarnings(\n",
    "    'ignore', '.*You seem to be using the pipelines sequentially on GPU.*')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data_dir = os.path.join(os.getcwd(), '..', 'data')\n",
    "data_dir = os.path.join(os.getcwd(), 'data')\n",
    "dirs_to_make = [\n",
    "    data_dir,\n",
    "    # os.path.join(data_dir, 'raw'),\n",
    "    # os.path.join(data_dir, 'processed'),\n",
    "]\n",
    "for d in dirs_to_make:\n",
    "    if not os.path.exists(d):\n",
    "        os.makedirs(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded protac.csv\n"
     ]
    }
   ],
   "source": [
    "protacdb_file = os.path.join(data_dir, 'PROTAC-DB.csv')\n",
    "protac_df = pd.read_csv(protacdb_file).reset_index(drop=True)\n",
    "\n",
    "protacdb_file = os.path.join(data_dir, 'PROTAC-DB-v2.csv')\n",
    "protac_v2_df = pd.read_csv(protacdb_file).reset_index(drop=True)\n",
    "\n",
    "print(f'Loaded protac.csv')\n",
    "\n",
    "old2new = {\n",
    "    'E3 ligase': 'E3 Ligase',\n",
    "}\n",
    "protac_df = protac_df.rename(columns=old2new)\n",
    "protac_v2_df = protac_v2_df.rename(columns=old2new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9380, 5388)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(protac_v2_df), len(protac_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PROTAC-DB\n",
      "Number of rows with all 3: 344\n",
      "Number of rows with Assay: 1008\n",
      "Number of rows with both DC50 and Dmax: 344\n",
      "Number of rows with DC50: 905\n",
      "Number of rows with Dmax: 726\n",
      "Number of rows with Percent degradation: 362\n",
      "\n",
      "PROTAC-DB-v2\n",
      "Number of rows with all 3: 909\n",
      "Number of rows with Assay: 1892\n",
      "Number of rows with both DC50 and Dmax: 909\n",
      "Number of rows with DC50: 1762\n",
      "Number of rows with Dmax: 1317\n",
      "Number of rows with Percent degradation: 1422\n"
     ]
    }
   ],
   "source": [
    "def print_dmax_dc_info(df):\n",
    "    num_all_notna = len(df.dropna(subset=['Assay (DC50/Dmax)', 'DC50 (nM)', 'Dmax (%)']).dropna(how='all').drop_duplicates())\n",
    "    num_assay_notna = len(df.dropna(subset=['Assay (DC50/Dmax)']).dropna(how='all').drop_duplicates())\n",
    "    num_both_notna = len(df.dropna(subset=['DC50 (nM)', 'Dmax (%)']).dropna(how='all').drop_duplicates())\n",
    "    num_dmax_notna = len(df.dropna(subset=['Dmax (%)']).dropna(how='all').drop_duplicates())\n",
    "    num_dc50_notna = len(df.dropna(subset=['DC50 (nM)']).dropna(how='all').drop_duplicates())\n",
    "    num_degr_notna = len(df.dropna(subset=['Percent degradation (%)']).dropna(how='all').drop_duplicates())\n",
    "    print(f'Number of rows with all 3: {num_all_notna}')\n",
    "    print(f'Number of rows with Assay: {num_assay_notna}')\n",
    "    print(f'Number of rows with both DC50 and Dmax: {num_both_notna}')\n",
    "    print(f'Number of rows with DC50: {num_dc50_notna}')\n",
    "    print(f'Number of rows with Dmax: {num_dmax_notna}')\n",
    "    print(f'Number of rows with Percent degradation: {num_degr_notna}')\n",
    "\n",
    "print('PROTAC-DB')\n",
    "print_dmax_dc_info(protac_df)\n",
    "print('')\n",
    "print('PROTAC-DB-v2')\n",
    "print_dmax_dc_info(protac_v2_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-100.0, -5.0, nan, 90.317, 1000.0, nan]\n",
      "[0.0]\n",
      "[96.0, 73.0]\n",
      "[1.0, 3.14]\n"
     ]
    }
   ],
   "source": [
    "def clean_string(s: str) -> str:\n",
    "    \"\"\" Clean a string by removing <, >, =, NaN, and ranges like 100-200.\n",
    "    Args:\n",
    "        s(str): string to clean\n",
    "    Returns:\n",
    "        str: cleaned string\n",
    "    \"\"\"\n",
    "    if pd.isnull(s) or s in {'nan', 'n/a', 'NaN', ''}:\n",
    "        return np.nan\n",
    "    if 'N.D.' in s:\n",
    "        return '0'\n",
    "    s = s.strip('(WB)').strip()\n",
    "    # # Combine regex operations for efficiency\n",
    "    # s = re.sub(r'[<=>]|NaN|[\\d]+[-~]', '', s)  # Remove <, >, =, NaN, and ranges like 100-200\n",
    "    # Remove <, >, =, NaN\n",
    "    s = re.sub(r'[<=>]|NaN', '', s)\n",
    "    # Replace ranges like 100-200 or 1~3 with the left-most value in the range\n",
    "    s = re.sub(r'\\b(\\d+)[-~]\\d+\\b', r'\\1', s)\n",
    "    # Replace (n/a) with nan\n",
    "    s = s.replace('(n/a)', 'nan')\n",
    "    s = re.sub(r'[~<=>% ]', '', s)  # Remove ~, <, >, =, % and spaces\n",
    "    return s\n",
    "\n",
    "\n",
    "def split_clean_str(s: str, return_floats: bool = False) -> Union[List[str], List[float]]:\n",
    "    \"\"\" Split a string by '/' and clean each part.\n",
    "    Args:\n",
    "        s(str): string to split\n",
    "        return_floats(bool): whether to return floats or strings\n",
    "    Returns:\n",
    "        list: list of cleaned strings or floats\n",
    "    \"\"\"\n",
    "    if pd.isnull(s) or s in {'nan', 'n/a', 'NaN', ''}:\n",
    "        return np.nan\n",
    "    cleaned_values = [clean_string(part.strip())\n",
    "                      for part in s.replace('(n/a)', 'nan').split('/')]\n",
    "    return [float(value) if return_floats else value for value in cleaned_values]\n",
    "\n",
    "\n",
    "print(split_clean_str('-100-200/-5/(n/a)/<=90.317/>1000/NaN', return_floats=True))\n",
    "print(split_clean_str('N.D.', return_floats=True))\n",
    "print(split_clean_str('96/73 (WB)', return_floats=True))\n",
    "print(split_clean_str('1.0~3/3.14', return_floats=True))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "813\n",
      "848\n"
     ]
    }
   ],
   "source": [
    "def get_assay_texts(df: pd.DataFrame, assay_column: str) -> List[str]:\n",
    "    tmp = df[assay_column].dropna()\n",
    "    if tmp.empty:\n",
    "        return []\n",
    "    return tmp.unique().tolist()\n",
    "\n",
    "\n",
    "def clean_assay_text(assay):\n",
    "    tmp = assay.replace('/', ' and ')\n",
    "    tmp = tmp.replace('BRD4 BD1 and 2', 'BRD4 BD1 and BRD4 BD2')\n",
    "    tmp = tmp.replace('(Ba and F3 WT)', '(Ba/F3 WT)')\n",
    "    tmp = tmp.replace('(EGFR L858R and T790M)', '(EGFR L858R/T790M)')\n",
    "    return tmp\n",
    "\n",
    "\n",
    "assays = {}\n",
    "for c in protac_df.columns:\n",
    "    if 'Assay' in c:\n",
    "        assays[c] = get_assay_texts(protac_df, c)\n",
    "texts = list(set([x for y in assays.values() for x in y]))\n",
    "print(len(texts))\n",
    "print(sum([len(x) for x in assays.values()]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_dc50_info(sentence):\n",
    "    # Regex patterns for proteins/genes, cell types, and treatment hours\n",
    "    protein_regex = r\"Degradation of total\\s(.+?)\\s(in|after|using|proteins)\"\n",
    "    cell_regex = r\"in\\s([A-Za-z0-9-/.;\\(\\)\\s\\+]+)\\scells\"\n",
    "    treatment_regex = r\"after\\s(\\d+/?\\d*?/?\\d*?\\s?h)\"\n",
    "\n",
    "    # Extracting protein information\n",
    "    if 'total' in sentence.lower():\n",
    "        protein_match = re.search(protein_regex, sentence)\n",
    "        proteins = protein_match.group(1).split(' and ') if protein_match else [\n",
    "            re.search(r\"Degradation of\\s([A-Za-z0-9-]+)\", sentence).group(1)]\n",
    "    else:\n",
    "        if ' in ' in sentence.lower():\n",
    "            proteins = sentence.split(' in ')[0].split('Degradation of ')[-1]\n",
    "            proteins = proteins.split('/') if '/' in proteins else [proteins]\n",
    "        else:\n",
    "            protein_match = re.search(protein_regex, sentence)\n",
    "            proteins = protein_match.group(1).split(\n",
    "                '/') if protein_match else [re.search(r\"Degradation of\\s([A-Za-z0-9-\\/]+)\", sentence).group(1)]\n",
    "    # Handle special cases...\n",
    "    if 'BRD4 short/long' in sentence:\n",
    "        proteins = ['BRD4 short', 'BRD4 long']\n",
    "    if 'BRD4 BD1/2' in sentence:\n",
    "        proteins = ['BRD4 BD1', 'BRD4 BD2']\n",
    "    elif 'BRD4 BD1' in sentence:\n",
    "        proteins = ['BRD4 BD1']\n",
    "    if 'EGFR L858R/T790M' in sentence:\n",
    "        proteins = ['EGFR L858R/T790M']\n",
    "    if 'EGFR del19/T790M/C797S' in sentence:\n",
    "        proteins = ['EGFR del19/T790M/C797S']\n",
    "\n",
    "    # Extracting cell types\n",
    "    cell_match = re.search(cell_regex, sentence)\n",
    "    cells = cell_match.group(1).split('/') if cell_match else np.nan\n",
    "    # Handle special cases...\n",
    "    if 'Ba/F3' in sentence:\n",
    "        # Replace any occurences that contain 'Ba' or 'F3' with 'Ba/F3' and remove duplicates while preserving the order in the other cells\n",
    "        cells = ['Ba/F3' if 'Ba' in c or 'F3' in c else c for c in cells]\n",
    "        cells.pop(cells.index('Ba/F3'))\n",
    "    if 'ER-positive breast cancer cell lines' in sentence:\n",
    "        cells = ['ER-positive breast cancer cell lines']\n",
    "    if 'LNCaP (AR T878A)' in sentence:\n",
    "        cells = ['LNCaP']\n",
    "    if 'in A152T neurons' in sentence:\n",
    "        cells = ['A152T neurons']\n",
    "    if 'of Rpn13 in MM.1S after' in sentence:\n",
    "        cells = ['MM.1S']\n",
    "    if 'Primary Cardiomyocytes' in sentence:\n",
    "        cells = ['Primary Cardiomyocytes']\n",
    "    if ' HDAC6 in MM1S after' in sentence:\n",
    "        cells = ['MM.1S']\n",
    "\n",
    "    # Extracting treatment hours\n",
    "    treatment_hours_match = re.search(treatment_regex, sentence)\n",
    "    if treatment_hours_match:\n",
    "        treatment_hours = treatment_hours_match.group(1).strip('h')\n",
    "        treatment_hours = split_clean_str(treatment_hours, return_floats=True)\n",
    "    else:\n",
    "        treatment_hours = np.nan\n",
    "\n",
    "    return {\n",
    "        'Target (Parsed)': proteins,\n",
    "        'Cell Type': cells,\n",
    "        'Treatment Time (h)': treatment_hours,\n",
    "    }\n",
    "\n",
    "\n",
    "corner_cases = [\n",
    "    # 'Degradation of BRD4',\n",
    "    # 'Degradation of BRD4 short/long in HeLa cells after 24 h treatment',\n",
    "    # 'Degradation of BRD4 BD1 assessed by EGFP/mCherry reporter assay',\n",
    "    # 'Degradation of BRD4 BD1/2 assessed by EGFP/mCherry reporter assay',\n",
    "    # 'Degradation of WT/Exon 20 Ins EGFR in OVCAR8/HeLa cells after 24 h treatment',\n",
    "    # 'Degradation of TPM3-TRKA/TRKA in KM12/HEL cells after 6 h treatment',\n",
    "    # 'Degradation of Exon 19 del/L858R EGFR in HCC827/H3255 cells after 24 h treatment',\n",
    "    # 'Degradation of NPM-ALK/EML4-ALK in SU-DHL-1/NCI-H2228 cells after 16 h treatment',\n",
    "    # 'Degradation of BCR-ABL T315I in Ba/F3 cells after 24 h treatment',\n",
    "    # 'Degradation of BCR-ABL T315I in MOL/(Ba/F3)/R4;11 cells after 24 h treatment',\n",
    "    # 'Degradation of ALK in H3122/Karpas 299/Kelly cells 16 h treatment',\n",
    "    'Degradation of AR in LNCaP/VCaP AR+ cells after 6 h treatment',\n",
    "    'Degradation of BRD4 BD1/2 assessed by EGFP/mCherry reporter assay',\n",
    "    'Degradation of BRD4 BD1 assessed by EGFP/mCherry reporter assay',\n",
    "    'Degradation of PARP1 in Primary Cardiomyocytes after 24 h treatment',\n",
    "    'Degradation of HDAC6 in MM1S after 6 h treatment by in-cell ELISA analysis',\n",
    "    'Degradation of total tau/P-tau in A152T neurons after 24 h treatment',\n",
    "    'Degradation of Rpn13 in MM.1S after 16 h treatment',\n",
    "    'Degradation of HDAC6 in MM1S after 6 h treatment by in-cell ELISA analysis',\n",
    "]\n",
    "\n",
    "# for assay in assays[\"Assay (DC50/Dmax)\"][-5:] + corner_cases:\n",
    "#     if len(assay) < 5:\n",
    "#         continue\n",
    "#     print(assay)\n",
    "#     extracted_info = extract_dc50_info(assay)\n",
    "#     proteins, cells, treatment_hours = extracted_info[\n",
    "#         'Target (Parsed)'], extracted_info['Cell Type'], extracted_info['Treatment Time (h)']\n",
    "#     print(proteins, \"|\", cells, \"|\", treatment_hours)\n",
    "#     print('-' * 80)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_dc50_dmax_df(df):\n",
    "    param_cols = ['DC50 (nM)', 'Dmax (%)']\n",
    "    dc50_dmax_df = df.dropna(subset=param_cols + [\"Assay (DC50/Dmax)\"], how='all')\n",
    "    dc50_dmax_df = dc50_dmax_df[dc50_dmax_df[\"Assay (DC50/Dmax)\"].notnull()]\n",
    "    return dc50_dmax_df.drop_duplicates()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The 'Dmax (%)' column in PROTAC-DB-v2 has two entries which are _dates_ (you never stop surprising me, PROTAC-DB). Convert them to NaNs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# If any entry in the 'Dmax (%)' column contains the character ':', then it is a\n",
    "# date and it needs to be set to NaN\n",
    "def clean_dmax(df):\n",
    "    df['Dmax (%)'] = df['Dmax (%)'].apply(lambda x: np.nan if ':' in str(x) else x)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c889fc12d4a040a78fbfdc506696ea9f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Extracting DC50/Dmax info:   0%|          | 0/1008 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Compound ID</th>\n",
       "      <th>Uniprot</th>\n",
       "      <th>Target</th>\n",
       "      <th>E3 Ligase</th>\n",
       "      <th>PDB</th>\n",
       "      <th>Name</th>\n",
       "      <th>Smiles</th>\n",
       "      <th>DC50 (nM)</th>\n",
       "      <th>Dmax (%)</th>\n",
       "      <th>Assay (DC50/Dmax)</th>\n",
       "      <th>...</th>\n",
       "      <th>Hydrogen Bond Acceptor Count</th>\n",
       "      <th>Hydrogen Bond Donor Count</th>\n",
       "      <th>Rotatable Bond Count</th>\n",
       "      <th>Topological Polar Surface Area</th>\n",
       "      <th>Molecular Formula</th>\n",
       "      <th>InChI</th>\n",
       "      <th>InChI Key</th>\n",
       "      <th>Target (Parsed)</th>\n",
       "      <th>Cell Type</th>\n",
       "      <th>Treatment Time (h)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>560.00</td>\n",
       "      <td>80.0</td>\n",
       "      <td>Degradation of BRD9 in HeLa cells after 4 h tr...</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>3</td>\n",
       "      <td>22</td>\n",
       "      <td>199.15</td>\n",
       "      <td>C54H69FN8O10S</td>\n",
       "      <td>InChI=1S/C54H69FN8O10S/c1-34-47(74-33-58-34)35...</td>\n",
       "      <td>MXAKQOVZPDLCDK-UDVNCTHFSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>HeLa</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>1.76</td>\n",
       "      <td>95.0</td>\n",
       "      <td>Degradation of BRD9 in RI-1 cells after 8 h tr...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>RI-1</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>4.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Degradation of HiBiT-BRD9 in HEK293 cells afte...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>HiBiT-BRD9</td>\n",
       "      <td>HEK293</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>2.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Degradation of BRD9 in EOL-1/A-204 cells after...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>EOL-1</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>8.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Degradation of BRD9 in EOL-1/A-204 cells after...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>A-204</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 92 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Compound ID Uniprot Target E3 Ligase  PDB   Name  \\\n",
       "0           11  Q9H8M2   BRD9       VHL  NaN    NaN   \n",
       "1           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "2           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "3           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "4           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "\n",
       "                                              Smiles  DC50 (nM)  Dmax (%)  \\\n",
       "0  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...     560.00      80.0   \n",
       "1  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       1.76      95.0   \n",
       "2  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       4.00       NaN   \n",
       "3  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       2.00       NaN   \n",
       "4  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       8.00       NaN   \n",
       "\n",
       "                                   Assay (DC50/Dmax)  ...  \\\n",
       "0  Degradation of BRD9 in HeLa cells after 4 h tr...  ...   \n",
       "1  Degradation of BRD9 in RI-1 cells after 8 h tr...  ...   \n",
       "2  Degradation of HiBiT-BRD9 in HEK293 cells afte...  ...   \n",
       "3  Degradation of BRD9 in EOL-1/A-204 cells after...  ...   \n",
       "4  Degradation of BRD9 in EOL-1/A-204 cells after...  ...   \n",
       "\n",
       "  Hydrogen Bond Acceptor Count Hydrogen Bond Donor Count Rotatable Bond Count  \\\n",
       "0                           16                         3                   22   \n",
       "1                           14                         3                   19   \n",
       "2                           14                         3                   19   \n",
       "3                           14                         3                   19   \n",
       "4                           14                         3                   19   \n",
       "\n",
       "  Topological Polar Surface Area Molecular Formula  \\\n",
       "0                         199.15     C54H69FN8O10S   \n",
       "1                         180.69      C53H67FN8O8S   \n",
       "2                         180.69      C53H67FN8O8S   \n",
       "3                         180.69      C53H67FN8O8S   \n",
       "4                         180.69      C53H67FN8O8S   \n",
       "\n",
       "                                               InChI  \\\n",
       "0  InChI=1S/C54H69FN8O10S/c1-34-47(74-33-58-34)35...   \n",
       "1  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "2  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "3  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "4  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "\n",
       "                     InChI Key Target (Parsed) Cell Type Treatment Time (h)  \n",
       "0  MXAKQOVZPDLCDK-UDVNCTHFSA-N            BRD9      HeLa                4.0  \n",
       "1  ZAGCLFXBHOXXEN-JPTLTNPLSA-N            BRD9      RI-1                8.0  \n",
       "2  ZAGCLFXBHOXXEN-JPTLTNPLSA-N      HiBiT-BRD9    HEK293               24.0  \n",
       "3  ZAGCLFXBHOXXEN-JPTLTNPLSA-N            BRD9     EOL-1               18.0  \n",
       "4  ZAGCLFXBHOXXEN-JPTLTNPLSA-N            BRD9     A-204               18.0  \n",
       "\n",
       "[5 rows x 92 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsed table len: 1205\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a788394f66594587b03025bd8f3d9c51",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Extracting DC50/Dmax info:   0%|          | 0/1892 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Compound ID</th>\n",
       "      <th>Uniprot</th>\n",
       "      <th>Target</th>\n",
       "      <th>E3 Ligase</th>\n",
       "      <th>PDB</th>\n",
       "      <th>Name</th>\n",
       "      <th>Smiles</th>\n",
       "      <th>DC50 (nM)</th>\n",
       "      <th>Dmax (%)</th>\n",
       "      <th>Assay (DC50/Dmax)</th>\n",
       "      <th>...</th>\n",
       "      <th>Hydrogen Bond Acceptor Count</th>\n",
       "      <th>Hydrogen Bond Donor Count</th>\n",
       "      <th>Rotatable Bond Count</th>\n",
       "      <th>Topological Polar Surface Area</th>\n",
       "      <th>Molecular Formula</th>\n",
       "      <th>InChI</th>\n",
       "      <th>InChI Key</th>\n",
       "      <th>Target (Parsed)</th>\n",
       "      <th>Cell Type</th>\n",
       "      <th>Treatment Time (h)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>560.00</td>\n",
       "      <td>80.0</td>\n",
       "      <td>Degradation of BRD9 in HeLa cells after 4 h tr...</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>3</td>\n",
       "      <td>22</td>\n",
       "      <td>199.15</td>\n",
       "      <td>C54H69FN8O10S</td>\n",
       "      <td>InChI=1S/C54H69FN8O10S/c1-34-47(74-33-58-34)35...</td>\n",
       "      <td>MXAKQOVZPDLCDK-UDVNCTHFSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>HeLa</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>1.76</td>\n",
       "      <td>95.0</td>\n",
       "      <td>Degradation of BRD9 in RI-1 cells after 8 h tr...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>RI-1</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>4.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Degradation of HiBiT-BRD9 in HEK293 cells afte...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>HiBiT-BRD9</td>\n",
       "      <td>HEK293</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>2.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Degradation of BRD9 in EOL-1/A-204 cells after...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>EOL-1</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>22</td>\n",
       "      <td>Q9H8M2</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>VZ185</td>\n",
       "      <td>COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...</td>\n",
       "      <td>8.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Degradation of BRD9 in EOL-1/A-204 cells after...</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>180.69</td>\n",
       "      <td>C53H67FN8O8S</td>\n",
       "      <td>InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...</td>\n",
       "      <td>ZAGCLFXBHOXXEN-JPTLTNPLSA-N</td>\n",
       "      <td>BRD9</td>\n",
       "      <td>A-204</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 92 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Compound ID Uniprot Target E3 Ligase  PDB   Name  \\\n",
       "0           11  Q9H8M2   BRD9       VHL  NaN    NaN   \n",
       "1           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "2           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "3           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "4           22  Q9H8M2   BRD9       VHL  NaN  VZ185   \n",
       "\n",
       "                                              Smiles  DC50 (nM)  Dmax (%)  \\\n",
       "0  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...     560.00      80.0   \n",
       "1  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       1.76      95.0   \n",
       "2  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       4.00       NaN   \n",
       "3  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       2.00       NaN   \n",
       "4  COC1=CC(C2=CN(C)C(=O)C3=CN=CC=C23)=CC(OC)=C1CN...       8.00       NaN   \n",
       "\n",
       "                                   Assay (DC50/Dmax)  ...  \\\n",
       "0  Degradation of BRD9 in HeLa cells after 4 h tr...  ...   \n",
       "1  Degradation of BRD9 in RI-1 cells after 8 h tr...  ...   \n",
       "2  Degradation of HiBiT-BRD9 in HEK293 cells afte...  ...   \n",
       "3  Degradation of BRD9 in EOL-1/A-204 cells after...  ...   \n",
       "4  Degradation of BRD9 in EOL-1/A-204 cells after...  ...   \n",
       "\n",
       "  Hydrogen Bond Acceptor Count Hydrogen Bond Donor Count Rotatable Bond Count  \\\n",
       "0                           16                         3                   22   \n",
       "1                           14                         3                   19   \n",
       "2                           14                         3                   19   \n",
       "3                           14                         3                   19   \n",
       "4                           14                         3                   19   \n",
       "\n",
       "  Topological Polar Surface Area Molecular Formula  \\\n",
       "0                         199.15     C54H69FN8O10S   \n",
       "1                         180.69      C53H67FN8O8S   \n",
       "2                         180.69      C53H67FN8O8S   \n",
       "3                         180.69      C53H67FN8O8S   \n",
       "4                         180.69      C53H67FN8O8S   \n",
       "\n",
       "                                               InChI  \\\n",
       "0  InChI=1S/C54H69FN8O10S/c1-34-47(74-33-58-34)35...   \n",
       "1  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "2  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "3  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "4  InChI=1S/C53H67FN8O8S/c1-33-46(71-32-57-33)34-...   \n",
       "\n",
       "                     InChI Key Target (Parsed) Cell Type Treatment Time (h)  \n",
       "0  MXAKQOVZPDLCDK-UDVNCTHFSA-N            BRD9      HeLa                4.0  \n",
       "1  ZAGCLFXBHOXXEN-JPTLTNPLSA-N            BRD9      RI-1                8.0  \n",
       "2  ZAGCLFXBHOXXEN-JPTLTNPLSA-N      HiBiT-BRD9    HEK293               24.0  \n",
       "3  ZAGCLFXBHOXXEN-JPTLTNPLSA-N            BRD9     EOL-1               18.0  \n",
       "4  ZAGCLFXBHOXXEN-JPTLTNPLSA-N            BRD9     A-204               18.0  \n",
       "\n",
       "[5 rows x 92 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsed table len: 2264\n"
     ]
    }
   ],
   "source": [
    "dfs = {}\n",
    "\n",
    "for name, df in [('protac-db', protac_df), ('protac-db-v2', protac_v2_df)]:\n",
    "    dc50_dmax_df = get_dc50_dmax_df(clean_dmax(df))\n",
    "\n",
    "    parsed_table = []\n",
    "    for i, row in tqdm(dc50_dmax_df.iterrows(), total=len(dc50_dmax_df), desc='Extracting DC50/Dmax info'):\n",
    "        assay = row[\"Assay (DC50/Dmax)\"]\n",
    "        if len(assay) < 5:\n",
    "            continue\n",
    "        extracted_info = extract_dc50_info(assay)\n",
    "        extracted_info['DC50 (nM)'] = split_clean_str(\n",
    "            row['DC50 (nM)'], return_floats=True)\n",
    "        extracted_info['Dmax (%)'] = split_clean_str(\n",
    "            row['Dmax (%)'], return_floats=True)\n",
    "\n",
    "        # Get the max len of each list in the extracted info\n",
    "        max_len = max([len(v)\n",
    "                    for v in extracted_info.values() if isinstance(v, list)])\n",
    "        for i in range(max_len):\n",
    "            row_tmp = row.copy().to_dict()\n",
    "            row_tmp.update({k: v[i % len(v)] if isinstance(v, list)\n",
    "                        else v for k, v in extracted_info.items()})\n",
    "            parsed_table.append(row_tmp)\n",
    "\n",
    "    parsed_table = pd.DataFrame(parsed_table)\n",
    "    display(parsed_table.head())\n",
    "    print(f'Parsed table len: {len(parsed_table)}')\n",
    "    dfs[name] = parsed_table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def canonize_smiles(smi):\n",
    "    return Chem.MolToSmiles(Chem.MolFromSmiles(smi))\n",
    "\n",
    "dfs['protac-db']['Smiles'] = dfs['protac-db']['Smiles'].apply(canonize_smiles)\n",
    "dfs['protac-db-v2']['Smiles'] = dfs['protac-db-v2']['Smiles'].apply(canonize_smiles)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of entries in protac-db: 1205\n",
      "Number of entries in protac-db-v2: 2264\n",
      "Number of shared entries: 1249\n",
      "Number of total entries: 2232\n"
     ]
    }
   ],
   "source": [
    "# Get the number of entries in both dfs\n",
    "print(f'Number of entries in protac-db: {len(dfs[\"protac-db\"])}')\n",
    "print(f'Number of entries in protac-db-v2: {len(dfs[\"protac-db-v2\"])}')\n",
    "# Get the number of entries shared between the two dfs\n",
    "predict_cols = [\"Smiles\", \"DC50 (nM)\", \"Dmax (%)\", \"E3 Ligase\", \"Uniprot\", \"Cell Type\"]\n",
    "print(f'Number of shared entries: {len(dfs[\"protac-db\"].merge(dfs[\"protac-db-v2\"], on=predict_cols, how=\"inner\"))}')\n",
    "# Get the number of total entries without duplicates\n",
    "print(f'Number of total entries: {len(dfs[\"protac-db\"].append(dfs[\"protac-db-v2\"]).drop_duplicates(subset=predict_cols))}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}