03_antibiotic_xgboost.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import XGBClassifier\n",
    "import pandas as pd\n",
    "from rdkit import Chem\n",
    "from descriptastorus.descriptors.DescriptorGenerator import MakeGenerator\n",
    "from tqdm.notebook import tqdm\n",
    "from sklearn.model_selection import train_test_split\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.utils import resample"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read the antibacterial data from the supporting material from Stokes et al into a dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "ab_df = pd.read_excel(\"mmc1.xlsx\",sheet_name=1,skiprows=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to be consistent with previous notebooks, set active and inactive to 1 and 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "ab_df['active'] = [int(x == \"Active\") for x in ab_df.Activity]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "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>Mean_Inhibition</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>Name</th>\n",
       "      <th>Activity</th>\n",
       "      <th>active</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.041572</td>\n",
       "      <td>Cc1cc(O)c(C(=O)NC(C(=O)NC2C(=O)N3C(C(=O)O)=C(C...</td>\n",
       "      <td>CEFPIRAMIDE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.041876</td>\n",
       "      <td>CON=C1CN(c2nc3c(cc2F)c(=O)c(C(=O)O)cn3C2CC2)CC...</td>\n",
       "      <td>GEMIFLOXACIN MESYLATE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.041916</td>\n",
       "      <td>CCC(C)CCCCC(=O)NC(CCN)C(=O)NC(C(=O)NC(CCN)C(=O...</td>\n",
       "      <td>POLYMYXIN B SULFATE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.041964</td>\n",
       "      <td>Cl.N=C(N)n1cccn1</td>\n",
       "      <td>PRAXADINE HYDROCHLORIDE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.042295</td>\n",
       "      <td>Cl.Cl.N=C(NCCCCCCNC(=N)NC(=N)Nc1ccc(Cl)cc1)NC(...</td>\n",
       "      <td>CHLORHEXIDINE DIHYDROCHLORIDE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Mean_Inhibition                                             SMILES                           Name Activity  active\n",
       "0         0.041572  Cc1cc(O)c(C(=O)NC(C(=O)NC2C(=O)N3C(C(=O)O)=C(C...                    CEFPIRAMIDE   Active       1\n",
       "1         0.041876  CON=C1CN(c2nc3c(cc2F)c(=O)c(C(=O)O)cn3C2CC2)CC...          GEMIFLOXACIN MESYLATE   Active       1\n",
       "2         0.041916  CCC(C)CCCCC(=O)NC(CCN)C(=O)NC(C(=O)NC(CCN)C(=O...            POLYMYXIN B SULFATE   Active       1\n",
       "3         0.041964                                   Cl.N=C(N)n1cccn1        PRAXADINE HYDROCHLORIDE   Active       1\n",
       "4         0.042295  Cl.Cl.N=C(NCCCCCCNC(=N)NC(=N)Nc1ccc(Cl)cc1)NC(...  CHLORHEXIDINE DIHYDROCHLORIDE   Active       1"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ab_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use Brian Kelley's descriptastorus to generate descritptors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator = MakeGenerator((\"RDKit2D\", \"Morgan3Counts\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "20d338c5de8f4afcab4e0fdf9022d972",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=2335.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "desc_list = [generator.process(x) for x in tqdm(ab_df.SMILES)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make sure all of our descriptor calculations worked.  Descriptastorus sets the first element in the returned list to True or False to inidcate the success of the calculations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True    2335\n",
       "dtype: int64"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "desc_ok = [x[0] for x in desc_list]\n",
    "pd.Series(desc_ok).value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks like everything worked, add the descriptors to the dataframe. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "ab_df['desc'] = [x[1:] for x in desc_list]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "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>Mean_Inhibition</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>Name</th>\n",
       "      <th>Activity</th>\n",
       "      <th>active</th>\n",
       "      <th>desc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.041572</td>\n",
       "      <td>Cc1cc(O)c(C(=O)NC(C(=O)NC2C(=O)N3C(C(=O)O)=C(C...</td>\n",
       "      <td>CEFPIRAMIDE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "      <td>[1.3395052635747295, 1610.1816625187853, 30.12...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.041876</td>\n",
       "      <td>CON=C1CN(c2nc3c(cc2F)c(=O)c(C(=O)O)cn3C2CC2)CC...</td>\n",
       "      <td>GEMIFLOXACIN MESYLATE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "      <td>[5.549998829416112e-07, 1265.508949285827, 24....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.041916</td>\n",
       "      <td>CCC(C)CCCCC(=O)NC(CCN)C(=O)NC(C(=O)NC(CCN)C(=O...</td>\n",
       "      <td>POLYMYXIN B SULFATE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "      <td>[7.761161663762498e-06, 2563.1541142939495, 68...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.041964</td>\n",
       "      <td>Cl.N=C(N)n1cccn1</td>\n",
       "      <td>PRAXADINE HYDROCHLORIDE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "      <td>[6.399999284000083e-07, 180.45752214691183, 5....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.042295</td>\n",
       "      <td>Cl.Cl.N=C(NCCCCCCNC(=N)NC(=N)Nc1ccc(Cl)cc1)NC(...</td>\n",
       "      <td>CHLORHEXIDINE DIHYDROCHLORIDE</td>\n",
       "      <td>Active</td>\n",
       "      <td>1</td>\n",
       "      <td>[6.1249911528881385e-06, 888.2990570824002, 24...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Mean_Inhibition                                             SMILES  ... active                                               desc\n",
       "0         0.041572  Cc1cc(O)c(C(=O)NC(C(=O)NC2C(=O)N3C(C(=O)O)=C(C...  ...      1  [1.3395052635747295, 1610.1816625187853, 30.12...\n",
       "1         0.041876  CON=C1CN(c2nc3c(cc2F)c(=O)c(C(=O)O)cn3C2CC2)CC...  ...      1  [5.549998829416112e-07, 1265.508949285827, 24....\n",
       "2         0.041916  CCC(C)CCCCC(=O)NC(CCN)C(=O)NC(C(=O)NC(CCN)C(=O...  ...      1  [7.761161663762498e-06, 2563.1541142939495, 68...\n",
       "3         0.041964                                   Cl.N=C(N)n1cccn1  ...      1  [6.399999284000083e-07, 180.45752214691183, 5....\n",
       "4         0.042295  Cl.Cl.N=C(NCCCCCCNC(=N)NC(=N)Nc1ccc(Cl)cc1)NC(...  ...      1  [6.1249911528881385e-06, 888.2990570824002, 24...\n",
       "\n",
       "[5 rows x 6 columns]"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ab_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create an XGBoost Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb = XGBClassifier()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's build quick model to see how well we predict from the internal data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "train,test = train_test_split(ab_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Build a model with the data. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.array([x for x in train.desc])\n",
    "y = train.active\n",
    "xgb.fit(x,y)\n",
    "test_x = np.array([x for x in test.desc])\n",
    "xgb_prob = xgb.predict_proba(test_x)\n",
    "xgb_pred = xgb.predict(test_x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make a boxplot of probabilities vs activity lables to ensure that we're looking at this the right way"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0.5, 'probability'), Text(0.5, 0, 'activity')]"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = sns.boxplot(xgb_pred,xgb_prob[:,1])\n",
    "ax.set(xlabel=\"activity\",ylabel=\"probability\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Calculate a ROC AUC for our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8683423099940156"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "roc_auc_score(test.active,xgb_prob[:,1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This looks pretty good, now let's build a model with all of the antibiotic data. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1, gamma=0,\n",
       "              learning_rate=0.1, max_delta_step=0, max_depth=3,\n",
       "              min_child_weight=1, missing=None, n_estimators=100, n_jobs=1,\n",
       "              nthread=None, objective='binary:logistic', random_state=0,\n",
       "              reg_alpha=0, reg_lambda=1, scale_pos_weight=1, seed=None,\n",
       "              silent=None, subsample=1, verbosity=1)"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb_full = XGBClassifier()\n",
    "x_full = np.array([x for x in ab_df.desc])\n",
    "y_full = ab_df.active\n",
    "xgb_full.fit(x_full,y_full)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have a model, we can use it to make som predictions on the Broad Institute's repurposing set.  First, we'll grab this set from the web. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "url = \"https://s3.amazonaws.com/data.clue.io/repurposing/downloads/repurposing_samples_20180907.txt\"\n",
    "repurposing_df = pd.read_csv(url, comment = \"!\", sep=\"\\t\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's take a look at the first few rows to see what's there. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "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>broad_id</th>\n",
       "      <th>pert_iname</th>\n",
       "      <th>qc_incompatible</th>\n",
       "      <th>purity</th>\n",
       "      <th>vendor</th>\n",
       "      <th>catalog_no</th>\n",
       "      <th>vendor_name</th>\n",
       "      <th>expected_mass</th>\n",
       "      <th>smiles</th>\n",
       "      <th>InChIKey</th>\n",
       "      <th>pubchem_cid</th>\n",
       "      <th>deprecated_broad_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BRD-K89787693-001-01-1</td>\n",
       "      <td>[sar9,met(o2)11]-substance-p</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Tocris</td>\n",
       "      <td>1178</td>\n",
       "      <td>2-({5-amino-2-[({1-[6-amino-2-({[1-(2-amino-5-...</td>\n",
       "      <td>1,392.73</td>\n",
       "      <td>CC(C)C[C@H](NC(=O)CN(C)C(=O)[C@H](Cc1ccccc1)NC...</td>\n",
       "      <td>OUPXSLGGCPUZJJ-SARDKLJWSA-N</td>\n",
       "      <td>163829.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BRD-K88956297-003-01-9</td>\n",
       "      <td>1-((Z)-3-Chloroallyl)-1,3,5,7-tetraazaadamanta...</td>\n",
       "      <td>0</td>\n",
       "      <td>94.41</td>\n",
       "      <td>BidePharm</td>\n",
       "      <td>BD51690</td>\n",
       "      <td>1-((Z)-3-Chloroallyl)-1,3,5,7-tetraazaadamanta...</td>\n",
       "      <td>215.106</td>\n",
       "      <td>Cl\\C=C/C[N+]12CN3CN(CN(C3)C1)C2</td>\n",
       "      <td>LDLCEGCJYSDJLX-UPHRSURJSA-N</td>\n",
       "      <td>5846454.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BRD-A86415025-050-01-0</td>\n",
       "      <td>1-(1,2-Diphenylethyl)piperidine-(+/-)</td>\n",
       "      <td>0</td>\n",
       "      <td>98.04</td>\n",
       "      <td>Tocris</td>\n",
       "      <td>360</td>\n",
       "      <td>(?)-1-(1,2-Diphenylethyl)piperidine maleate</td>\n",
       "      <td>265.183</td>\n",
       "      <td>C(C(N1CCCCC1)c1ccccc1)c1ccccc1</td>\n",
       "      <td>JQWJJJYHVHNXJH-UHFFFAOYSA-N</td>\n",
       "      <td>206666.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BRD-A95802703-001-01-0</td>\n",
       "      <td>1-(2-chloro-5-methylphenoxy)-3-(isopropylamino...</td>\n",
       "      <td>0</td>\n",
       "      <td>92.38</td>\n",
       "      <td>Enamine</td>\n",
       "      <td>Z1672746675</td>\n",
       "      <td>1-(2-chloro-5-methylphenoxy)-3-(isopropylamino...</td>\n",
       "      <td>257.118</td>\n",
       "      <td>CC(C)NCC(O)COc1cc(C)ccc1Cl</td>\n",
       "      <td>NJEIOWSBPCZKTL-UHFFFAOYSA-N</td>\n",
       "      <td>20497006.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>BRD-K96055017-003-02-7</td>\n",
       "      <td>1-acetyl-4-methylpiperazine</td>\n",
       "      <td>0</td>\n",
       "      <td>90.28</td>\n",
       "      <td>Tocris</td>\n",
       "      <td>351</td>\n",
       "      <td>1-Acetyl-4-methylpiperazine hydrochloride</td>\n",
       "      <td>142.111</td>\n",
       "      <td>CN1CCN(CC1)C(C)=O</td>\n",
       "      <td>YSDBJKNOEWSFGA-UHFFFAOYSA-N</td>\n",
       "      <td>550758.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 broad_id                                         pert_iname  ...  pubchem_cid  deprecated_broad_id\n",
       "0  BRD-K89787693-001-01-1                       [sar9,met(o2)11]-substance-p  ...     163829.0                  NaN\n",
       "1  BRD-K88956297-003-01-9  1-((Z)-3-Chloroallyl)-1,3,5,7-tetraazaadamanta...  ...    5846454.0                  NaN\n",
       "2  BRD-A86415025-050-01-0              1-(1,2-Diphenylethyl)piperidine-(+/-)  ...     206666.0                  NaN\n",
       "3  BRD-A95802703-001-01-0  1-(2-chloro-5-methylphenoxy)-3-(isopropylamino...  ...   20497006.0                  NaN\n",
       "4  BRD-K96055017-003-02-7                        1-acetyl-4-methylpiperazine  ...     550758.0                  NaN\n",
       "\n",
       "[5 rows x 12 columns]"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "repurposing_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we saw in the previous notebook, the repurposing data contains duplicates and null values.  Let's get rid of these."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "repurposing_df.drop_duplicates('smiles',inplace=True)\n",
    "repurposing_df.dropna(axis=0,subset=['smiles'],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6244, 12)"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "repurposing_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As above, use descriptastorus to calcualte descriptors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7bf6cf9c645a4756a129ba067aaaa39b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=6244.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "repurposing_desc_list = [generator.process(x) for x in tqdm(repurposing_df.smiles)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True    6244\n",
       "dtype: int64"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "repurposing_desc_ok = [x[0] for x in repurposing_desc_list]\n",
    "pd.Series(repurposing_desc_ok).value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "repurposing_df['desc'] = [x[1:] for x in repurposing_desc_list]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "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>broad_id</th>\n",
       "      <th>pert_iname</th>\n",
       "      <th>qc_incompatible</th>\n",
       "      <th>purity</th>\n",
       "      <th>vendor</th>\n",
       "      <th>catalog_no</th>\n",
       "      <th>vendor_name</th>\n",
       "      <th>expected_mass</th>\n",
       "      <th>smiles</th>\n",
       "      <th>InChIKey</th>\n",
       "      <th>pubchem_cid</th>\n",
       "      <th>deprecated_broad_id</th>\n",
       "      <th>desc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BRD-K89787693-001-01-1</td>\n",
       "      <td>[sar9,met(o2)11]-substance-p</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Tocris</td>\n",
       "      <td>1178</td>\n",
       "      <td>2-({5-amino-2-[({1-[6-amino-2-({[1-(2-amino-5-...</td>\n",
       "      <td>1,392.73</td>\n",
       "      <td>CC(C)C[C@H](NC(=O)CN(C)C(=O)[C@H](Cc1ccccc1)NC...</td>\n",
       "      <td>OUPXSLGGCPUZJJ-SARDKLJWSA-N</td>\n",
       "      <td>163829.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[1.5818916260184153, 3220.1773701867137, 72.81...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BRD-K88956297-003-01-9</td>\n",
       "      <td>1-((Z)-3-Chloroallyl)-1,3,5,7-tetraazaadamanta...</td>\n",
       "      <td>0</td>\n",
       "      <td>94.41</td>\n",
       "      <td>BidePharm</td>\n",
       "      <td>BD51690</td>\n",
       "      <td>1-((Z)-3-Chloroallyl)-1,3,5,7-tetraazaadamanta...</td>\n",
       "      <td>215.106</td>\n",
       "      <td>Cl\\C=C/C[N+]12CN3CN(CN(C3)C1)C2</td>\n",
       "      <td>LDLCEGCJYSDJLX-UPHRSURJSA-N</td>\n",
       "      <td>5846454.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[1.8009625314016358, 230.2533564975087, 9.5960...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BRD-A86415025-050-01-0</td>\n",
       "      <td>1-(1,2-Diphenylethyl)piperidine-(+/-)</td>\n",
       "      <td>0</td>\n",
       "      <td>98.04</td>\n",
       "      <td>Tocris</td>\n",
       "      <td>360</td>\n",
       "      <td>(?)-1-(1,2-Diphenylethyl)piperidine maleate</td>\n",
       "      <td>265.183</td>\n",
       "      <td>C(C(N1CCCCC1)c1ccccc1)c1ccccc1</td>\n",
       "      <td>JQWJJJYHVHNXJH-UHFFFAOYSA-N</td>\n",
       "      <td>206666.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[1.8288898030917446, 499.804163233686, 13.6231...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BRD-A95802703-001-01-0</td>\n",
       "      <td>1-(2-chloro-5-methylphenoxy)-3-(isopropylamino...</td>\n",
       "      <td>0</td>\n",
       "      <td>92.38</td>\n",
       "      <td>Enamine</td>\n",
       "      <td>Z1672746675</td>\n",
       "      <td>1-(2-chloro-5-methylphenoxy)-3-(isopropylamino...</td>\n",
       "      <td>257.118</td>\n",
       "      <td>CC(C)NCC(O)COc1cc(C)ccc1Cl</td>\n",
       "      <td>NJEIOWSBPCZKTL-UHFFFAOYSA-N</td>\n",
       "      <td>20497006.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[2.41254152698502, 355.2727408671211, 12.83649...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>BRD-K96055017-003-02-7</td>\n",
       "      <td>1-acetyl-4-methylpiperazine</td>\n",
       "      <td>0</td>\n",
       "      <td>90.28</td>\n",
       "      <td>Tocris</td>\n",
       "      <td>351</td>\n",
       "      <td>1-Acetyl-4-methylpiperazine hydrochloride</td>\n",
       "      <td>142.111</td>\n",
       "      <td>CN1CCN(CC1)C(C)=O</td>\n",
       "      <td>YSDBJKNOEWSFGA-UHFFFAOYSA-N</td>\n",
       "      <td>550758.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[2.328221491209792, 127.5677964944704, 7.56047...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 broad_id  ...                                               desc\n",
       "0  BRD-K89787693-001-01-1  ...  [1.5818916260184153, 3220.1773701867137, 72.81...\n",
       "1  BRD-K88956297-003-01-9  ...  [1.8009625314016358, 230.2533564975087, 9.5960...\n",
       "2  BRD-A86415025-050-01-0  ...  [1.8288898030917446, 499.804163233686, 13.6231...\n",
       "3  BRD-A95802703-001-01-0  ...  [2.41254152698502, 355.2727408671211, 12.83649...\n",
       "4  BRD-K96055017-003-02-7  ...  [2.328221491209792, 127.5677964944704, 7.56047...\n",
       "\n",
       "[5 rows x 13 columns]"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "repurposing_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Format the descriptors for model building"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "repurposing_x = np.array([x for x in repurposing_df.desc])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Predict activity for the repurposing set and put this into the dataframe. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "repurposing_pred = xgb_full.predict_proba(repurposing_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "repurposing_df['pred'] = repurposing_pred[:,1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sort the dataframe by the predicted probabilities. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "repurposing_df.sort_values(\"pred\",ascending=False,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "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>broad_id</th>\n",
       "      <th>pert_iname</th>\n",
       "      <th>qc_incompatible</th>\n",
       "      <th>purity</th>\n",
       "      <th>vendor</th>\n",
       "      <th>catalog_no</th>\n",
       "      <th>vendor_name</th>\n",
       "      <th>expected_mass</th>\n",
       "      <th>smiles</th>\n",
       "      <th>InChIKey</th>\n",
       "      <th>pubchem_cid</th>\n",
       "      <th>deprecated_broad_id</th>\n",
       "      <th>desc</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4129</th>\n",
       "      <td>BRD-A74980173-213-01-1</td>\n",
       "      <td>gatifloxacin</td>\n",
       "      <td>0</td>\n",
       "      <td>77.22</td>\n",
       "      <td>MicroSource</td>\n",
       "      <td>1504272</td>\n",
       "      <td>GATIFLOXACIN</td>\n",
       "      <td>375.159</td>\n",
       "      <td>COc1c(N2CCNC(C)C2)c(F)cc2c1n(cc(C(O)=O)c2=O)C1CC1</td>\n",
       "      <td>XUBOMFCQGDBHNK-UHFFFAOYSA-N</td>\n",
       "      <td>4872633.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[2.0312285638536896, 983.3041997140859, 19.292...</td>\n",
       "      <td>0.976365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8106</th>\n",
       "      <td>BRD-K62374253-001-07-8</td>\n",
       "      <td>rufloxacin</td>\n",
       "      <td>0</td>\n",
       "      <td>97.36</td>\n",
       "      <td>MicroSource</td>\n",
       "      <td>1502335</td>\n",
       "      <td>RUFLOXACIN HYDROCHLORIDE</td>\n",
       "      <td>363.105</td>\n",
       "      <td>CN1CCN(CC1)c1c(F)cc2c3c1SCCn3cc(C(O)=O)c2=O</td>\n",
       "      <td>NJCJBUHJQLFDSW-UHFFFAOYSA-N</td>\n",
       "      <td>4979271.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[1.9743414651959867, 935.1055962233459, 17.714...</td>\n",
       "      <td>0.967965</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6632</th>\n",
       "      <td>BRD-A24228527-213-01-9</td>\n",
       "      <td>ofloxacin</td>\n",
       "      <td>0</td>\n",
       "      <td>98.67</td>\n",
       "      <td>MicroSource</td>\n",
       "      <td>1502044</td>\n",
       "      <td>OFLOXACIN</td>\n",
       "      <td>361.144</td>\n",
       "      <td>CC1COc2c(N3CCN(C)CC3)c(F)cc3c2n1cc(C(O)=O)c3=O</td>\n",
       "      <td>GSDSWSVVBLHKDQ-UHFFFAOYSA-N</td>\n",
       "      <td>3288597.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[2.004088915400979, 963.2881482992597, 18.5850...</td>\n",
       "      <td>0.966582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5246</th>\n",
       "      <td>BRD-K09471561-001-18-2</td>\n",
       "      <td>levofloxacin</td>\n",
       "      <td>0</td>\n",
       "      <td>97.70</td>\n",
       "      <td>MicroSource</td>\n",
       "      <td>1504260</td>\n",
       "      <td>LEVOFLOXACIN</td>\n",
       "      <td>361.144</td>\n",
       "      <td>C[C@H]1COc2c(N3CCN(C)CC3)c(F)cc3c2n1cc(C(O)=O)...</td>\n",
       "      <td>GSDSWSVVBLHKDQ-JTQLQIEISA-N</td>\n",
       "      <td>149096.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[2.004088915400979, 963.2881482992597, 18.5850...</td>\n",
       "      <td>0.966582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3468</th>\n",
       "      <td>BRD-K76534306-001-22-7</td>\n",
       "      <td>enrofloxacin</td>\n",
       "      <td>0</td>\n",
       "      <td>96.46</td>\n",
       "      <td>MicroSource</td>\n",
       "      <td>1503721</td>\n",
       "      <td>ENROFLOXACIN</td>\n",
       "      <td>359.165</td>\n",
       "      <td>CCN1CCN(CC1)c1cc2n(cc(C(O)=O)c(=O)c2cc1F)C1CC1</td>\n",
       "      <td>SPFYMRJSYKOXGV-UHFFFAOYSA-N</td>\n",
       "      <td>6957639.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[1.8342383024563755, 927.751757198637, 18.4219...</td>\n",
       "      <td>0.958741</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    broad_id    pert_iname  qc_incompatible  ...  deprecated_broad_id                                               desc      pred\n",
       "4129  BRD-A74980173-213-01-1  gatifloxacin                0  ...                  NaN  [2.0312285638536896, 983.3041997140859, 19.292...  0.976365\n",
       "8106  BRD-K62374253-001-07-8    rufloxacin                0  ...                  NaN  [1.9743414651959867, 935.1055962233459, 17.714...  0.967965\n",
       "6632  BRD-A24228527-213-01-9     ofloxacin                0  ...                  NaN  [2.004088915400979, 963.2881482992597, 18.5850...  0.966582\n",
       "5246  BRD-K09471561-001-18-2  levofloxacin                0  ...                  NaN  [2.004088915400979, 963.2881482992597, 18.5850...  0.966582\n",
       "3468  BRD-K76534306-001-22-7  enrofloxacin                0  ...                  NaN  [1.8342383024563755, 927.751757198637, 18.4219...  0.958741\n",
       "\n",
       "[5 rows x 14 columns]"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "repurposing_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a rank order column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "repurposing_df['rank'] = range(0,repurposing_df.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Add a molecule column to the dataframe so that we can do a substructure search. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9ed4b5cc466d4e81ab1923992c7638af",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=6244.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "repurposing_df['mol'] = [Chem.MolFromSmiles(x) for x in tqdm(repurposing_df.smiles)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a query molecule from Halicin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "halicin = Chem.MolFromSmarts(\"Nc1nnc(Sc2ncc(s2)[N+]([O-])=O)s1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<rdkit.Chem.rdchem.Mol at 0x7f2300a1d8a0>"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "halicin"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Do the substructure search and show the matching row. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0f97bd2bff6f41df80e8c1860efe950d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=6244.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\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>broad_id</th>\n",
       "      <th>pert_iname</th>\n",
       "      <th>qc_incompatible</th>\n",
       "      <th>purity</th>\n",
       "      <th>vendor</th>\n",
       "      <th>catalog_no</th>\n",
       "      <th>vendor_name</th>\n",
       "      <th>expected_mass</th>\n",
       "      <th>smiles</th>\n",
       "      <th>InChIKey</th>\n",
       "      <th>pubchem_cid</th>\n",
       "      <th>deprecated_broad_id</th>\n",
       "      <th>desc</th>\n",
       "      <th>pred</th>\n",
       "      <th>rank</th>\n",
       "      <th>mol</th>\n",
       "      <th>match</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8722</th>\n",
       "      <td>BRD-K32730239-001-01-0</td>\n",
       "      <td>SU3327</td>\n",
       "      <td>0</td>\n",
       "      <td>94.58</td>\n",
       "      <td>Tocris</td>\n",
       "      <td>3607</td>\n",
       "      <td>SU 3327</td>\n",
       "      <td>260.945</td>\n",
       "      <td>Nc1nnc(Sc2ncc(s2)[N+]([O-])=O)s1</td>\n",
       "      <td>NQQBNZBOOHHVQP-UHFFFAOYSA-N</td>\n",
       "      <td>11837140.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[2.1404489309233523, 494.32877134531054, 10.83...</td>\n",
       "      <td>0.046852</td>\n",
       "      <td>764</td>\n",
       "      <td>&lt;rdkit.Chem.rdchem.Mol object at 0x7f22c9c956c0&gt;</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    broad_id pert_iname  qc_incompatible  purity  ...      pred rank                                               mol match\n",
       "8722  BRD-K32730239-001-01-0     SU3327                0   94.58  ...  0.046852  764  <rdkit.Chem.rdchem.Mol object at 0x7f22c9c956c0>  True\n",
       "\n",
       "[1 rows x 17 columns]"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "repurposing_df['match'] = [x.HasSubstructMatch(halicin) for x in tqdm(repurposing_df.mol)]\n",
    "repurposing_df.query(\"match\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this case, Halicin ranks 765, much lower than what we saw with ChemProp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rebalance_data(df_in):\n",
    "    df_active = df_in.query(\"active == 1\")\n",
    "    df_inactive = df_in.query(\"active == 0\")\n",
    "    df_active_upsample = resample(df_active,replace=True,n_samples=df_inactive.shape[0])\n",
    "    df_upsample = df_inactive.append(df_active_upsample)\n",
    "    return df_upsample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Environment (conda_chemprop)",
   "language": "python",
   "name": "conda_chemprop"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}