{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import logging\n",
    "\n",
    "sys.path.append(\"..\")\n",
    "\n",
    "# Set logging level to warning\n",
    "logging.basicConfig(level=logging.WARNING)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.10/site-packages/xgboost/compat.py:36: FutureWarning: pandas.Int64Index is deprecated and will be removed from pandas in a future version. Use pandas.Index with the appropriate dtype instead.\n",
      "  from pandas import MultiIndex, Int64Index\n"
     ]
    }
   ],
   "source": [
    "import protac_degradation_predictor as pdp\n",
    "\n",
    "import pandas as pd\n",
    "from tqdm.notebook import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "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>Uniprot</th>\n",
       "      <th>Cell Line Identifier</th>\n",
       "      <th>Smiles</th>\n",
       "      <th>E3 Ligase</th>\n",
       "      <th>DC50 (nM)</th>\n",
       "      <th>Dmax (%)</th>\n",
       "      <th>Active</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>53.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2136</th>\n",
       "      <td>O60885</td>\n",
       "      <td>HEK293</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>63.1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2137</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCCOCCO...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>125.9</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2138</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCC(=O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>158.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2139</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCC(=O)N[C...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>31.6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2140</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCC(=O)N[C@H...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>39.8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2141 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Uniprot Cell Line Identifier  \\\n",
       "0     Q07817               MOLT-4   \n",
       "1     Q07817               MOLT-4   \n",
       "2     Q07817               MOLT-4   \n",
       "3     Q07817               MOLT-4   \n",
       "4     Q07817               MOLT-4   \n",
       "...      ...                  ...   \n",
       "2136  O60885               HEK293   \n",
       "2137  Q05397            A549 Cas9   \n",
       "2138  Q05397            A549 Cas9   \n",
       "2139  Q05397            A549 Cas9   \n",
       "2140  Q05397            A549 Cas9   \n",
       "\n",
       "                                                 Smiles E3 Ligase  DC50 (nM)  \\\n",
       "0     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "1     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "2     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "3     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "4     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL       53.0   \n",
       "...                                                 ...       ...        ...   \n",
       "2136  Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL       63.1   \n",
       "2137  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCCOCCO...       VHL      125.9   \n",
       "2138  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCC(=O)...       VHL      158.5   \n",
       "2139  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCC(=O)N[C...       VHL       31.6   \n",
       "2140  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCC(=O)N[C@H...       VHL       39.8   \n",
       "\n",
       "      Dmax (%) Active  \n",
       "0          NaN    NaN  \n",
       "1          NaN    NaN  \n",
       "2          NaN    NaN  \n",
       "3          NaN    NaN  \n",
       "4        100.0   True  \n",
       "...        ...    ...  \n",
       "2136       NaN    NaN  \n",
       "2137       NaN    NaN  \n",
       "2138       NaN    NaN  \n",
       "2139       NaN    NaN  \n",
       "2140       NaN    NaN  \n",
       "\n",
       "[2141 rows x 7 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_dir = \"data\"\n",
    "protac_df = pd.read_csv(os.path.join(data_dir, \"PROTAC-Degradation-DB.csv\"))\n",
    "\n",
    "pDC50_threshold = 6.0\n",
    "Dmax_threshold = 0.6\n",
    "\n",
    "# Map E3 Ligase Iap to IAP\n",
    "protac_df['E3 Ligase'] = protac_df['E3 Ligase'].str.replace('Iap', 'IAP')\n",
    "protac_df['Active'] = protac_df.apply(\n",
    "    lambda x: pdp.is_active(x['DC50 (nM)'], x['Dmax (%)'], pDC50_threshold=pDC50_threshold, Dmax_threshold=Dmax_threshold), axis=1\n",
    ")\n",
    "protac_df[['Uniprot', 'Cell Line Identifier', 'Smiles', 'E3 Ligase', 'DC50 (nM)', 'Dmax (%)', 'Active']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "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>Uniprot</th>\n",
       "      <th>Cell Line Identifier</th>\n",
       "      <th>Smiles</th>\n",
       "      <th>E3 Ligase</th>\n",
       "      <th>DC50 (nM)</th>\n",
       "      <th>Dmax (%)</th>\n",
       "      <th>Active</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2136</th>\n",
       "      <td>O60885</td>\n",
       "      <td>HEK293</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>63.1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2137</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCCOCCO...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>125.9</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2138</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCC(=O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>158.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2139</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCC(=O)N[C...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>31.6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2140</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCC(=O)N[C@H...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>39.8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1284 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Uniprot Cell Line Identifier  \\\n",
       "0     Q07817               MOLT-4   \n",
       "1     Q07817               MOLT-4   \n",
       "2     Q07817               MOLT-4   \n",
       "3     Q07817               MOLT-4   \n",
       "5     Q07817               MOLT-4   \n",
       "...      ...                  ...   \n",
       "2136  O60885               HEK293   \n",
       "2137  Q05397            A549 Cas9   \n",
       "2138  Q05397            A549 Cas9   \n",
       "2139  Q05397            A549 Cas9   \n",
       "2140  Q05397            A549 Cas9   \n",
       "\n",
       "                                                 Smiles E3 Ligase  DC50 (nM)  \\\n",
       "0     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "1     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "2     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "3     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "5     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL        NaN   \n",
       "...                                                 ...       ...        ...   \n",
       "2136  Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL       63.1   \n",
       "2137  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCCOCCO...       VHL      125.9   \n",
       "2138  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCC(=O)...       VHL      158.5   \n",
       "2139  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCC(=O)N[C...       VHL       31.6   \n",
       "2140  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCC(=O)N[C@H...       VHL       39.8   \n",
       "\n",
       "      Dmax (%) Active  \n",
       "0          NaN    NaN  \n",
       "1          NaN    NaN  \n",
       "2          NaN    NaN  \n",
       "3          NaN    NaN  \n",
       "5          NaN    NaN  \n",
       "...        ...    ...  \n",
       "2136       NaN    NaN  \n",
       "2137       NaN    NaN  \n",
       "2138       NaN    NaN  \n",
       "2139       NaN    NaN  \n",
       "2140       NaN    NaN  \n",
       "\n",
       "[1284 rows x 7 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nan_df = protac_df[protac_df['Active'].isna()]\n",
    "nan_df[['Uniprot', 'Cell Line Identifier', 'Smiles', 'E3 Ligase', 'DC50 (nM)', 'Dmax (%)', 'Active']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "81b5e5ce4f944e2e8f10a3bae30708e7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1284 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The target protein P31750 is not in the database. Using the default target protein.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The target protein A8DG50 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein A8DG50 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein A8DG50 is not in the database. Using the default target protein.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The target protein P00520 is not in the database. Using the default target protein.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n",
      "WARNING:root:The cell line nan is not in the database. Using the default cell line.\n"
     ]
    },
    {
     "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>Uniprot</th>\n",
       "      <th>Cell Line Identifier</th>\n",
       "      <th>Smiles</th>\n",
       "      <th>E3 Ligase</th>\n",
       "      <th>mean_active_prob</th>\n",
       "      <th>majority_vote_active</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.661548</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.664419</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.671933</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.662499</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Q07817</td>\n",
       "      <td>MOLT-4</td>\n",
       "      <td>Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.649880</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2136</th>\n",
       "      <td>O60885</td>\n",
       "      <td>HEK293</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.596114</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2137</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCCOCCO...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.652442</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2138</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCC(=O)...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.645811</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2139</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCC(=O)N[C...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.667549</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2140</th>\n",
       "      <td>Q05397</td>\n",
       "      <td>A549 Cas9</td>\n",
       "      <td>CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCC(=O)N[C@H...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.654636</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1284 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Uniprot Cell Line Identifier  \\\n",
       "0     Q07817               MOLT-4   \n",
       "1     Q07817               MOLT-4   \n",
       "2     Q07817               MOLT-4   \n",
       "3     Q07817               MOLT-4   \n",
       "5     Q07817               MOLT-4   \n",
       "...      ...                  ...   \n",
       "2136  O60885               HEK293   \n",
       "2137  Q05397            A549 Cas9   \n",
       "2138  Q05397            A549 Cas9   \n",
       "2139  Q05397            A549 Cas9   \n",
       "2140  Q05397            A549 Cas9   \n",
       "\n",
       "                                                 Smiles E3 Ligase  \\\n",
       "0     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL   \n",
       "1     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL   \n",
       "2     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL   \n",
       "3     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL   \n",
       "5     Cc1ncsc1-c1ccc([C@H](C)NC(=O)[C@@H]2C[C@@H](O)...       VHL   \n",
       "...                                                 ...       ...   \n",
       "2136  Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL   \n",
       "2137  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCCOCCO...       VHL   \n",
       "2138  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCCOCC(=O)...       VHL   \n",
       "2139  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCOCC(=O)N[C...       VHL   \n",
       "2140  CNC(=O)c1ccccc1Nc1cc(Nc2ccc(N3CCN(CCC(=O)N[C@H...       VHL   \n",
       "\n",
       "      mean_active_prob  majority_vote_active  \n",
       "0             0.661548                  True  \n",
       "1             0.664419                  True  \n",
       "2             0.671933                  True  \n",
       "3             0.662499                  True  \n",
       "5             0.649880                  True  \n",
       "...                ...                   ...  \n",
       "2136          0.596114                  True  \n",
       "2137          0.652442                  True  \n",
       "2138          0.645811                  True  \n",
       "2139          0.667549                  True  \n",
       "2140          0.654636                  True  \n",
       "\n",
       "[1284 rows x 6 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from collections import defaultdict\n",
    "\n",
    "predicted_df = []\n",
    "# Iterate over batched rows in the PROTAC dataframe (use tqdm for progress bar)\n",
    "batch = defaultdict(list)\n",
    "batch_size = 10\n",
    "for index, row in tqdm(nan_df.iterrows(), total=nan_df.shape[0]):\n",
    "    if len(batch) < batch_size:\n",
    "        batch['protac_smiles'].append(row['Smiles'])\n",
    "        batch['e3_ligase'].append(row['E3 Ligase'])\n",
    "        batch['target_uniprot'].append(row['Uniprot'])\n",
    "        batch['cell_line'].append(row['Cell Line Identifier'])\n",
    "    else:\n",
    "        active_protac_prob = pdp.get_protac_active_proba(\n",
    "            batch['protac_smiles'],\n",
    "            batch['e3_ligase'],\n",
    "            batch['target_uniprot'],\n",
    "            batch['cell_line'],\n",
    "            device='cpu', # Default to 'cpu'\n",
    "        )\n",
    "        for i in range(len(batch['protac_smiles'])):\n",
    "            mean_active_protac_prob = active_protac_prob['mean'][i]\n",
    "            majority_vote_protac_prob = active_protac_prob['majority_vote'][i]\n",
    "            row['mean_active_prob'] = mean_active_protac_prob\n",
    "            row['majority_vote_active'] = majority_vote_protac_prob\n",
    "            predicted_df.append(row.copy())\n",
    "        batch = defaultdict(list)\n",
    "\n",
    "\n",
    "for index, row in tqdm(nan_df.iterrows(), total=nan_df.shape[0]):\n",
    "    protac_smiles = row['Smiles']\n",
    "    e3_ligase = row['E3 Ligase']\n",
    "    target_uniprot = row['Uniprot']\n",
    "    cell_line = row['Cell Line Identifier']\n",
    "\n",
    "    try:\n",
    "        active_protac_prob = pdp.get_protac_active_proba(\n",
    "            protac_smiles,\n",
    "            e3_ligase,\n",
    "            target_uniprot,\n",
    "            cell_line,\n",
    "            device='cpu', # Default to 'cpu'\n",
    "        )\n",
    "        mean_active_protac_prob = active_protac_prob['mean']\n",
    "        majority_vote_protac_prob = active_protac_prob['majority_vote']\n",
    "    except Exception as e:\n",
    "        logging.error(f\"Error: {e}\")\n",
    "        mean_active_protac_prob = float('nan')\n",
    "        majority_vote_protac_prob = float('nan')\n",
    "    \n",
    "    row['mean_active_prob'] = mean_active_protac_prob\n",
    "    row['majority_vote_active'] = majority_vote_protac_prob\n",
    "    predicted_df.append(row.copy())\n",
    "\n",
    "predicted_df = pd.DataFrame(predicted_df)\n",
    "predicted_df[['Uniprot', 'Cell Line Identifier', 'Smiles', 'E3 Ligase', 'mean_active_prob', 'majority_vote_active']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    1284.000000\n",
       "mean        0.456657\n",
       "std         0.163657\n",
       "min         0.034244\n",
       "25%         0.348055\n",
       "50%         0.493245\n",
       "75%         0.577750\n",
       "max         0.930988\n",
       "Name: mean_active_prob, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "count      1284\n",
       "unique        2\n",
       "top       False\n",
       "freq        665\n",
       "Name: majority_vote_active, dtype: object"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get mean active probability and majority vote active statistics\n",
    "mean_active_prob_stats = predicted_df['mean_active_prob'].describe()\n",
    "majority_vote_active_stats = predicted_df['majority_vote_active'].describe()\n",
    "\n",
    "display(mean_active_prob_stats)\n",
    "display(majority_vote_active_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the distribution of mean active probability and majority vote active probability\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(1, 2, figsize=(12, 6))\n",
    "\n",
    "sns.histplot(predicted_df['mean_active_prob'], bins=50, ax=ax[0], kde=True)\n",
    "ax[0].set_title(\"Mean Active Probability Distribution\")\n",
    "ax[0].set_xlabel(\"Mean Active Probability\")\n",
    "ax[0].set_ylabel(\"Frequency\")\n",
    "# Plot the mean of the mean active probability as a vertical line\n",
    "ax[0].axvline(predicted_df['mean_active_prob'].mean(), color='r', linestyle='--', label=f\"Mean: {predicted_df['mean_active_prob'].mean():.2f}\")\n",
    "\n",
    "sns.histplot(predicted_df['majority_vote_active'], bins=2, ax=ax[1])\n",
    "ax[1].set_title(\"Majority Vote Active Probability Distribution\")\n",
    "ax[1].set_xlabel(\"Majority Vote Active Probability\")\n",
    "ax[1].set_ylabel(\"Frequency\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "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>Dmax (%)</th>\n",
       "      <th>pDC50</th>\n",
       "      <th>Uniprot</th>\n",
       "      <th>Cell Line Identifier</th>\n",
       "      <th>Smiles</th>\n",
       "      <th>E3 Ligase</th>\n",
       "      <th>mean_active_prob</th>\n",
       "      <th>majority_vote_active</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>Q9UHD2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.881667</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>Q9UHD2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.862472</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>Q9UHD2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.854805</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>Q9UHD2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.853683</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>Q9UHD2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.844864</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1865</th>\n",
       "      <td>61.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q13547</td>\n",
       "      <td>HCT116-53BPI(+/-)</td>\n",
       "      <td>Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.153555</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.920819</td>\n",
       "      <td>P40763</td>\n",
       "      <td>MOLM-16</td>\n",
       "      <td>NC(=O)CC[C@H](NC(=O)[C@@H]1CC[C@@H]2CCN(C(=O)C...</td>\n",
       "      <td>CRBN</td>\n",
       "      <td>0.140929</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020</th>\n",
       "      <td>NaN</td>\n",
       "      <td>7.769551</td>\n",
       "      <td>P40763</td>\n",
       "      <td>SU-DHL-1</td>\n",
       "      <td>NC(=O)CC[C@H](NC(=O)[C@@H]1CC[C@@H]2CCN(C(=O)C...</td>\n",
       "      <td>CRBN</td>\n",
       "      <td>0.135795</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2096</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.153663</td>\n",
       "      <td>Q9BY41</td>\n",
       "      <td>Jurkat</td>\n",
       "      <td>O=C1CCC(N2C(=O)c3cccc(NCCCCCCCCCCCNC(=O)c4cccc...</td>\n",
       "      <td>CRBN</td>\n",
       "      <td>0.162395</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2116</th>\n",
       "      <td>NaN</td>\n",
       "      <td>8.105130</td>\n",
       "      <td>P10636</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CC(=O)N1CCCN(c2nc(NCCc3cccs3)nc(N(C)CC(=O)NCCO...</td>\n",
       "      <td>VHL</td>\n",
       "      <td>0.141287</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>154 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Dmax (%)     pDC50 Uniprot Cell Line Identifier  \\\n",
       "143        NaN  6.000000  Q9UHD2                  NaN   \n",
       "144        NaN  6.000000  Q9UHD2                  NaN   \n",
       "145        NaN  6.000000  Q9UHD2                  NaN   \n",
       "146        NaN  6.000000  Q9UHD2                  NaN   \n",
       "147        NaN  6.000000  Q9UHD2                  NaN   \n",
       "...        ...       ...     ...                  ...   \n",
       "1865      61.0       NaN  Q13547    HCT116-53BPI(+/-)   \n",
       "2019       NaN  6.920819  P40763              MOLM-16   \n",
       "2020       NaN  7.769551  P40763             SU-DHL-1   \n",
       "2096       NaN  6.153663  Q9BY41               Jurkat   \n",
       "2116       NaN  8.105130  P10636                  NaN   \n",
       "\n",
       "                                                 Smiles E3 Ligase  \\\n",
       "143   Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL   \n",
       "144   Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL   \n",
       "145   Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL   \n",
       "146   Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL   \n",
       "147   Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL   \n",
       "...                                                 ...       ...   \n",
       "1865  Cc1ncsc1-c1ccc(CNC(=O)[C@@H]2C[C@@H](O)CN2C(=O...       VHL   \n",
       "2019  NC(=O)CC[C@H](NC(=O)[C@@H]1CC[C@@H]2CCN(C(=O)C...      CRBN   \n",
       "2020  NC(=O)CC[C@H](NC(=O)[C@@H]1CC[C@@H]2CCN(C(=O)C...      CRBN   \n",
       "2096  O=C1CCC(N2C(=O)c3cccc(NCCCCCCCCCCCNC(=O)c4cccc...      CRBN   \n",
       "2116  CC(=O)N1CCCN(c2nc(NCCc3cccs3)nc(N(C)CC(=O)NCCO...       VHL   \n",
       "\n",
       "      mean_active_prob  majority_vote_active  \n",
       "143           0.881667                  True  \n",
       "144           0.862472                  True  \n",
       "145           0.854805                  True  \n",
       "146           0.853683                  True  \n",
       "147           0.844864                  True  \n",
       "...                ...                   ...  \n",
       "1865          0.153555                 False  \n",
       "2019          0.140929                 False  \n",
       "2020          0.135795                 False  \n",
       "2096          0.162395                 False  \n",
       "2116          0.141287                 False  \n",
       "\n",
       "[154 rows x 8 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "confidence_threshold = 0.8\n",
    "\n",
    "tmp = predicted_df[(predicted_df['mean_active_prob'] > confidence_threshold) | (predicted_df['mean_active_prob'] < (1-confidence_threshold))].copy()\n",
    "tmp['pDC50'] = tmp['DC50 (nM)'].apply(lambda x: -np.log10(x*1e-9) if x > 0 else float('nan'))\n",
    "display(tmp[['Dmax (%)', 'pDC50', 'Uniprot', 'Cell Line Identifier', 'Smiles', 'E3 Ligase', 'mean_active_prob', 'majority_vote_active']])\n",
    "\n",
    "# Plot a scatter plot of mean active probability vs. Dmax (%)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(8, 6))\n",
    "\n",
    "sns.scatterplot(data=tmp, x='pDC50', y='mean_active_prob', hue='E3 Ligase', ax=ax)\n",
    "ax.set_title(\"Mean Active Probability vs. Dmax (%)\")\n",
    "ax.set_xlabel(\"pDC50\")\n",
    "ax.set_ylabel(\"Mean Active Probability\")\n",
    "\n",
    "ax.grid(axis='both', linestyle='-', alpha=0.5)\n",
    "\n",
    "plt.ylim(0, 1)\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Plot a scatter plot of mean active probability vs. Dmax (%)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(8, 6))\n",
    "\n",
    "sns.scatterplot(data=tmp, x='Dmax (%)', y='mean_active_prob', hue='E3 Ligase', ax=ax)\n",
    "ax.set_title(\"Mean Active Probability vs. Dmax (%)\")\n",
    "ax.set_xlabel(\"Dmax (%)\")\n",
    "ax.set_ylabel(\"Mean Active Probability\")\n",
    "\n",
    "ax.grid(axis='both', linestyle='-', alpha=0.5)\n",
    "\n",
    "plt.ylim(0, 1)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[VHL]\tThe PROTAC is active with probability {'mean': 0.5639503, 'majority_vote': True}\n",
      "[CRBN]\tThe PROTAC is active with probability {'mean': 0.55152416, 'majority_vote': True}\n",
      "[DCAF11]\tThe PROTAC is inactive with probability {'mean': 0.34527245, 'majority_vote': False}\n",
      "[DCAF15]\tThe PROTAC is inactive with probability {'mean': 0.39864042, 'majority_vote': False}\n",
      "[DCAF16]\tThe PROTAC is inactive with probability {'mean': 0.38463345, 'majority_vote': False}\n",
      "[MDM2]\tThe PROTAC is inactive with probability {'mean': 0.3296026, 'majority_vote': False}\n",
      "[Mdm2]\tThe PROTAC is inactive with probability {'mean': 0.3296026, 'majority_vote': False}\n",
      "[XIAP]\tThe PROTAC is inactive with probability {'mean': 0.3460035, 'majority_vote': False}\n",
      "[cIAP1]\tThe PROTAC is inactive with probability {'mean': 0.28262994, 'majority_vote': False}\n",
      "[IAP]\tThe PROTAC is inactive with probability {'mean': 0.3460035, 'majority_vote': False}\n",
      "[Iap]\tThe PROTAC is inactive with probability {'mean': 0.3460035, 'majority_vote': False}\n",
      "[AhR]\tThe PROTAC is inactive with probability {'mean': 0.37744805, 'majority_vote': False}\n",
      "[RNF4]\tThe PROTAC is inactive with probability {'mean': 0.32840127, 'majority_vote': False}\n",
      "[RNF114]\tThe PROTAC is inactive with probability {'mean': 0.33398724, 'majority_vote': False}\n",
      "[FEM1B]\tThe PROTAC is inactive with probability {'mean': 0.31875828, 'majority_vote': False}\n",
      "[Ubr1]\tThe PROTAC is inactive with probability {'mean': 0.44158635, 'majority_vote': False}\n"
     ]
    }
   ],
   "source": [
    "e3_ligases = list(pdp.config.config.e3_ligase2uniprot.keys())\n",
    "for e3_ligase in e3_ligases:\n",
    "    active_protac = pdp.is_protac_active(\n",
    "        protac_smiles,\n",
    "        e3_ligase,\n",
    "        target_uniprot,\n",
    "        cell_line,\n",
    "        device='cpu', # Default to 'cpu'\n",
    "        proba_threshold=0.5, # Default value\n",
    "    )\n",
    "    active_protac_prob = pdp.get_protac_active_proba(\n",
    "        protac_smiles,\n",
    "        e3_ligase,\n",
    "        target_uniprot,\n",
    "        cell_line,\n",
    "        device='cpu', # Default to 'cpu'\n",
    "    )\n",
    "    print(f'[{e3_ligase}]\\tThe PROTAC is {\"active\" if active_protac else \"inactive\"} with probability {active_protac_prob}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "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
}