14_BSEP_modeling.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "extensions": {
      "jupyter_dashboards": {
        "activeView": "report_default",
        "version": 1,
        "views": {
          "grid_default": {
            "cellMargin": 10,
            "defaultCellHeight": 20,
            "maxColumns": 12,
            "name": "grid",
            "type": "grid"
          },
          "report_default": {
            "name": "report",
            "type": "report"
          }
        }
      }
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.6.12"
    },
    "colab": {
      "name": "14_BSEP-modeling.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ravichas/AMPL-Tutorial/blob/master/14_BSEP_modeling.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0V2ybLgAH-0V"
      },
      "source": [
        "# BSEP Inhibition Model Creation and Prediction\n",
        "The data source is https://github.com/ATOMconsortium/AMPL/tree/master/atomsci/ddm/examples/BSEP.  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FYelWdOkIGwC"
      },
      "source": [
        "The notebook is based on the modeling procedure published in the following manscript, https://pubmed.ncbi.nlm.nih.gov/33502191/, and provides AMPL code to create/test Random Forest (RF) model. The notebook ends by testing the RF model using an external test data. Please note the datasets used in this notebook came from AMPL github site.  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7J41ky8ULQdh"
      },
      "source": [
        "# Basic biology of BSEP gene\n",
        "You can read about BSEP gene here, \n",
        "\n",
        "https://en.wikipedia.org/wiki/ABCB11\n",
        "\n",
        "and/or here,\n",
        "\n",
        "http://useast.ensembl.org/Homo_sapiens/Gene/Summary?db=core;g=ENSG00000276582;r=CHR_HSCHR2_1_CTG7_2:168939927-168953422 \n",
        "\n",
        "BSEP or ABCB11 is a gene associated with the disease progressive familial intrahepatic cholestasis type 2 (PFIC2).PFIC2 is shown to be caused by mutations in the ABCB11 gene which increases the risk of hepatocellular carcinoma in early life.\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VyhYeRqMxO_3"
      },
      "source": [
        "The notebook will run through the following steps: \n",
        "\n",
        "1.   Clone the AMPL git repository and install the software  \n",
        "2.   Use a data file from AMPL repository as input\n",
        "3.   Split the data \n",
        "4.   Train a model\n",
        "5.   Show how you can make predictions on the model.   \n",
        "6.   If you want to use your input data, please make sure your input data should be in CSV format that contains the following information: \n",
        "        *   Row of column headers\n",
        "        *   contain a column of SMILES strings for the compounds you want to run predictions on.\n",
        "        * Optional information: \n",
        "                *   A column of unique compound IDs; if none is provided, one will be generated for you\n",
        "                *   A column of binary activity values, if you have measured IC50s for the input compounds: 0 if the IC50 < 100 uM, 1 if IC50 >= 100 uM."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VMHN84qZXZNO"
      },
      "source": [
        "## Here is a sample input dataset with four componds"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JIxydMlpdWxS"
      },
      "source": [
        "|base_rdkit_smiles | compound_name| pIC50 |relation|active |\n",
        "|------------------|------------|-------|--------|-------|\n",
        "|CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C3C(=O)c4c(O)cccc4[C@@](C)(O)C3CC12\"\t|CHEMBL198\t|3.000000\t|<\t|0 |\n",
        "|C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]43C)C1C[C@@H](O)[C@]2(O)C(=O)CO|CHEMBL3039453\t|3.627825\t|NaN\t|1 |\n",
        "|Nc1nc2cc(Cl)ccc2o1\t|CHEMBL472566\t|3.000000\t|< |0|\n",
        "|COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]5CO[C@@H](C)O[C@H]5[C@H](O)[C@H]3O)[C@H]3COC(=O)[C@H]23)OCO4)cc(OC)c1O\t|CHEMBL44657\t|3.876148\t|<|\t0 |\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "od_Bn9W5gCh9"
      },
      "source": [
        "## Before you begin, make sure you close all other COLAB notebooks. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2DWqm4Cxm5h_"
      },
      "source": [
        "# Change Runtime settings\n",
        "\n",
        "## Plesae change your runtime settings to use GPU and high-memory, if you have \n",
        "\n",
        "## Runtime --> Change Runtime Type --> GPU"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rlYz7j65MDcb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4f986207-c756-4189-dafb-3464c8ded1ed"
      },
      "source": [
        "!date # starting time"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Wed Aug 11 14:41:11 UTC 2021\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HxuncCafRD2q"
      },
      "source": [
        "# AMPL Installation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_rNx7PY_S4ew"
      },
      "source": [
        "! pip install rdkit-pypi\n",
        "! pip install --pre deepchem\n",
        "\n",
        "import deepchem\n",
        "# print(deepchem.__version__)\n",
        "! pip install umap\n",
        "! pip install llvmlite==0.35.0  --ignore-installed\n",
        "! pip install umap-learn\n",
        "! pip install molvs\n",
        "! pip install bravado"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GTCkT2vpt4LP"
      },
      "source": [
        "import deepchem as dc\n",
        "\n",
        "# get the Install AMPL_GPU_test.sh\n",
        "!wget https://raw.githubusercontent.com/ravichas/AMPL-Tutorial/master/config/install_AMPL_GPU_test.sh\n",
        "\n",
        "# run the script to install AMPL\n",
        "! chmod u+x install_AMPL_GPU_test.sh\n",
        "! ./install_AMPL_GPU_test.sh"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-BT3RB2wviMd"
      },
      "source": [
        "# Obtain the files from GitHub\n",
        "! wget https://raw.githubusercontent.com/ATOMconsortium/AMPL/master/atomsci/ddm/examples/BSEP/data/small_test_data.csv\n",
        "! wget https://raw.githubusercontent.com/ATOMconsortium/AMPL/master/atomsci/ddm/examples/BSEP/data/morgan_warner_combined_bsep_data.csv"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wtT7wqY0mXmX"
      },
      "source": [
        "# Load the libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jkE19oitxcFG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "96a6b68c-0103-492f-a7f1-40e230878ab3"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import numpy as np\n",
        "sns.set_context(\"poster\")\n",
        "sns.set_style(\"whitegrid\")\n",
        "sns.set_palette(\"Set2\")\n",
        "\n",
        "import pandas as pd\n",
        "import os, json, sys, glob, pickle\n",
        "\n",
        "from atomsci.ddm.pipeline import model_pipeline as mp\n",
        "from atomsci.ddm.pipeline import parameter_parser as parse\n",
        "from atomsci.ddm.pipeline import perf_data"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2021-08-11 14:42:59,072 Model tracker client not supported in your environment; will save models in filesystem only.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TECUt-ajmbkb"
      },
      "source": [
        "# Load the Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oYE_k-gty67Y"
      },
      "source": [
        "df = pd.read_csv(\"morgan_warner_combined_bsep_data.csv\")"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 142
        },
        "id": "2msnpfXmzDsR",
        "outputId": "2c0133d8-91d6-40cf-9580-ffbd76d9b419"
      },
      "source": [
        "df.head(3)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>compound_name</th>\n",
              "      <th>source</th>\n",
              "      <th>base_rdkit_smiles</th>\n",
              "      <th>relation</th>\n",
              "      <th>pIC50</th>\n",
              "      <th>active</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>fenoximone</td>\n",
              "      <td>Morgan</td>\n",
              "      <td>CSc1ccc(C(=O)c2[nH]c(=O)[nH]c2C)cc1</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>3.876148</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>OXEDRINE</td>\n",
              "      <td>Warner</td>\n",
              "      <td>CNCC(O)c1ccc(O)cc1</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>ccl4</td>\n",
              "      <td>Morgan</td>\n",
              "      <td>ClC(Cl)(Cl)Cl</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>3.876148</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  compound_name  source  ...     pIC50 active\n",
              "0    fenoximone  Morgan  ...  3.876148      0\n",
              "1      OXEDRINE  Warner  ...  3.000000      0\n",
              "2          ccl4  Morgan  ...  3.876148      0\n",
              "\n",
              "[3 rows x 6 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-1jKkKXOmsUl"
      },
      "source": [
        "## Display some molecules"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 901
        },
        "id": "_iGdSjKXctjv",
        "outputId": "e002102c-dfc5-453b-c6e4-cf5580b98ce7"
      },
      "source": [
        "from rdkit import Chem\n",
        "from rdkit.Chem import AllChem\n",
        "from rdkit.Chem import Draw\n",
        "from rdkit.Chem import rdDepictor\n",
        "from itertools import islice\n",
        "\n",
        "#Needed to show molecules\n",
        "from rdkit.Chem.Draw import IPythonConsole \n",
        "\n",
        "from rdkit.Chem.Draw.MolDrawing import MolDrawing, DrawingOptions \n",
        "from rdkit.Chem import PandasTools\n",
        "from rdkit.Chem.Draw import IPythonConsole\n",
        "\n",
        "\n",
        "molecules = [Chem.MolFromSmiles(smiles)\n",
        "             for smiles in islice(df['base_rdkit_smiles'],15)]\n",
        "\n",
        "Draw.MolsToGridImage(molecules,  subImgSize=(450,250))"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4tyzWOcBmzTg"
      },
      "source": [
        "# Modeling: Training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eHnYylOmxxQU",
        "outputId": "be04f86d-b937-4e86-c220-80b2925c1238"
      },
      "source": [
        "response_col = \"active\"\n",
        "compound_id = \"compound_name\"\n",
        "smiles_col = \"base_rdkit_smiles\"\n",
        "\n",
        "params = {\n",
        "        \"verbose\": \"True\",\n",
        "        \"system\": \"LC\",\n",
        "        \"lc_account\": 'None',\n",
        "        \"datastore\": \"False\",\n",
        "        \"save_results\": \"False\",\n",
        "        \"data_owner\": \"username\",\n",
        "        \"prediction_type\": \"classification\",\n",
        "        \"dataset_key\": \"/content/morgan_warner_combined_bsep_data.csv\",\n",
        "        \"id_col\": compound_id,\n",
        "        \"smiles_col\": smiles_col,\n",
        "        \"response_cols\": response_col,\n",
        "        \"previously_split\": \"False\",\n",
        "        \"split_only\": \"False\",\n",
        "        \"featurizer\": \"ecfp\",\n",
        "        \"model_type\": \"RF\",\n",
        "        \"verbose\": \"True\",\n",
        "        \"transformers\": \"True\",\n",
        "        'max_epochs': '70',\n",
        "        \"rerun\": \"False\",\n",
        "        \"result_dir\": \"/content\"\n",
        "        # \"ampl_version\": \" \"\n",
        "    }\n",
        "\n",
        "ampl_param = parse.wrapper(params)\n",
        "pl = mp.ModelPipeline(ampl_param)\n",
        "pl.train_model()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2021-08-11 14:43:59,365 Attempting to load featurized dataset\n",
            "2021-08-11 14:43:59,372 Exception when trying to load featurized data:\n",
            "DynamicFeaturization doesn't support get_featurized_dset_name()\n",
            "2021-08-11 14:43:59,373 Featurized dataset not previously saved for dataset morgan_warner_combined_bsep_data, creating new\n",
            "2021-08-11 14:43:59,383 Featurizing sample 0\n",
            "2021-08-11 14:44:03,117 Splitting data by scaffold\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "number of features: 1024\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "2021-08-11 14:44:03,439 Dataset split table saved to /content/morgan_warner_combined_bsep_data_train_valid_test_scaffold_3e6294c3-9e02-463c-a899-10adfca2357a.csv\n",
            "2021-08-11 14:44:03,463 Fitting random forest model\n",
            "2021-08-11 14:44:05,181 Fold 0: training roc_auc_score = 1.000, validation roc_auc_score = 0.795, test roc_auc_score = 0.747\n",
            "2021-08-11 14:44:05,530 Wrote model metadata to file /content/morgan_warner_combined_bsep_data/RF_ecfp_scaffold_classification/c6623e4c-b43c-404d-9bd7-59b86fa138bc/model_metadata.json\n",
            "2021-08-11 14:44:05,595 Wrote model metrics to file /content/morgan_warner_combined_bsep_data/RF_ecfp_scaffold_classification/c6623e4c-b43c-404d-9bd7-59b86fa138bc/model_metrics.json\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Wrote model tarball to /content/morgan_warner_combined_bsep_data_model_c6623e4c-b43c-404d-9bd7-59b86fa138bc.tar.gz\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h6zieHQJSGiv"
      },
      "source": [
        "# Explore the model tar and metadata files "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K47a88Gya95C",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4be7df08-5469-4072-eb8a-6d32f13282ee"
      },
      "source": [
        "!tar -tf *.tar.gz"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "./best_model/\n",
            "./best_model/model.joblib\n",
            "./model_metadata.json\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PXcTKu4Gc50z"
      },
      "source": [
        "!tar xzf *.tar.gz -C /tmp"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eRcW-Y1ohH71",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "64a96e3c-330a-4aeb-d78c-e57e6a1e2a47"
      },
      "source": [
        "!cat /tmp/model_metadata.json | head"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{\n",
            "    \"ecfp_specific\": {\n",
            "        \"ecfp_radius\": 2,\n",
            "        \"ecfp_size\": 1024\n",
            "    },\n",
            "    \"model_parameters\": {\n",
            "        \"ampl_version\": \"1.2.0\",\n",
            "        \"featurizer\": \"ecfp\",\n",
            "        \"hyperparam_uuid\": null,\n",
            "        \"model_bucket\": \"public\",\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JjbADzw15u9C",
        "outputId": "c7a133ea-bf0d-4e1b-a035-41eaa02a37c3"
      },
      "source": [
        "import joblib\n",
        "# load the model from disk\n",
        "loaded_model = joblib.load(\"/tmp/best_model/model.joblib\")\n",
        "loaded_model"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
              "                       criterion='gini', max_depth=None, max_features=32,\n",
              "                       max_leaf_nodes=None, max_samples=None,\n",
              "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                       min_samples_leaf=1, min_samples_split=2,\n",
              "                       min_weight_fraction_leaf=0.0, n_estimators=500,\n",
              "                       n_jobs=-1, oob_score=False, random_state=None, verbose=0,\n",
              "                       warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JUrEbaxteRmi",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "666e821e-e14d-44b7-d6bf-341b239b4557"
      },
      "source": [
        "import json\n",
        "import pandas as pd\n",
        "data = json.load(open('/tmp/model_metadata.json'))\n",
        "# pd.DataFrame(data[\"model_parameters\"])\n",
        "data['model_parameters']"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'ampl_version': '1.2.0',\n",
              " 'featurizer': 'ecfp',\n",
              " 'hyperparam_uuid': None,\n",
              " 'model_bucket': 'public',\n",
              " 'model_choice_score_type': 'roc_auc',\n",
              " 'model_type': 'RF',\n",
              " 'num_model_tasks': 1,\n",
              " 'prediction_type': 'classification',\n",
              " 'save_results': False,\n",
              " 'system': 'LC',\n",
              " 'time_generated': 1628693045.5109706,\n",
              " 'transformer_bucket': 'public',\n",
              " 'transformer_key': None,\n",
              " 'transformer_oid': None,\n",
              " 'transformers': True,\n",
              " 'uncertainty': True}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wsuaH3EaSKEa"
      },
      "source": [
        "# Make a prediction on the validation test"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QDmeonI6AQTq",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b513686b-4ac0-4e60-ce05-b5752f620e4a"
      },
      "source": [
        "pred_data = pl.model_wrapper.get_perf_data(subset=\"valid\", epoch_label=\"best\")\n",
        "pred_results = pred_data.get_prediction_results()\n",
        "pred_results\n",
        "print(f\"AUC score of validation set: {pred_results['prc_auc_score']:.3f}\")"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "AUC score of validation set: 0.743\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A-7cboqQEmZz"
      },
      "source": [
        "pred_data = pl.model_wrapper.get_perf_data(subset=\"test\", epoch_label=\"best\")\n",
        "pred_results = pred_data.get_prediction_results()\n",
        "# print(f\"R2 score of validation set: {pred_results['r2_score']:.3f}\")"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NjGuDF46EpDj",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "94bf57a5-7434-4a06-d621-84aabc867a5c"
      },
      "source": [
        "pred_results"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy_score': 0.696969696969697,\n",
              " 'bal_accuracy': 0.6948529411764706,\n",
              " 'confusion_matrix': [[[26, 8], [12, 20]]],\n",
              " 'cross_entropy': 0.5918038827402821,\n",
              " 'kappa': 0.3911439114391144,\n",
              " 'matthews_cc': 0.39407655228132893,\n",
              " 'npv': 0.6842105263157895,\n",
              " 'num_compounds': 66,\n",
              " 'prc_auc_score': 0.7296276022625086,\n",
              " 'precision': 0.7142857142857143,\n",
              " 'recall_score': 0.625,\n",
              " 'roc_auc_score': 0.7472426470588235}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ARMFa7QroT_H"
      },
      "source": [
        "# Use an external data set for prediction\n",
        "If you want to use your input data, please make sure your input data should be in CSV format that contains the following information:\n",
        "Row of column headers\n",
        "contain a column of SMILES strings for the compounds you want to run predictions on.\n",
        "Optional information:\n",
        "  *   A column of unique compound IDs; if none is provided, one will be generated for you\n",
        "  *   A column of binary activity values, if you have measured IC50s for the input compounds: 0 if the IC50 < 100 uM, 1 if IC50 >= 100 uM.\n",
        "\n",
        "Let us read a sample test data and make predictions.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pIslyPkKIYrr",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 173
        },
        "outputId": "bec6a99c-a25d-4079-ef31-135d727adb9b"
      },
      "source": [
        "test_data = pd.read_csv(\"small_test_data.csv\")\n",
        "test_data.head(4)"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>base_rdkit_smiles</th>\n",
              "      <th>compound_id</th>\n",
              "      <th>pIC50</th>\n",
              "      <th>relation</th>\n",
              "      <th>active</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...</td>\n",
              "      <td>CHEMBL198</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]...</td>\n",
              "      <td>CHEMBL3039453</td>\n",
              "      <td>3.627825</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Nc1nc2cc(Cl)ccc2o1</td>\n",
              "      <td>CHEMBL472566</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...</td>\n",
              "      <td>CHEMBL44657</td>\n",
              "      <td>3.876148</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                   base_rdkit_smiles  ... active\n",
              "0  CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...  ...      0\n",
              "1  C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]...  ...      1\n",
              "2                                 Nc1nc2cc(Cl)ccc2o1  ...      0\n",
              "3  COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...  ...      0\n",
              "\n",
              "[4 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p4IaisQqJAt7"
      },
      "source": [
        "# ?pl.predict_on_smiles\n",
        "list_cpds = list(test_data.base_rdkit_smiles.values)"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iFg5Juy_JuX-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "ae800be9-d199-4ba9-c99c-4ac84e4adc41"
      },
      "source": [
        "test_data.rename(columns={\"compound_id\": \"compound_name\"}, inplace=True)\n",
        "test_data.head(5)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>base_rdkit_smiles</th>\n",
              "      <th>compound_name</th>\n",
              "      <th>pIC50</th>\n",
              "      <th>relation</th>\n",
              "      <th>active</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...</td>\n",
              "      <td>CHEMBL198</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]...</td>\n",
              "      <td>CHEMBL3039453</td>\n",
              "      <td>3.627825</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Nc1nc2cc(Cl)ccc2o1</td>\n",
              "      <td>CHEMBL472566</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...</td>\n",
              "      <td>CHEMBL44657</td>\n",
              "      <td>3.876148</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>CC(=O)[C@H]1[C@H](C#N)C[C@H]2[C@@H]3CC=C4CC(O)...</td>\n",
              "      <td>CHEMBL4063685</td>\n",
              "      <td>3.876148</td>\n",
              "      <td>&lt;</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                   base_rdkit_smiles  ... active\n",
              "0  CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...  ...      0\n",
              "1  C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]...  ...      1\n",
              "2                                 Nc1nc2cc(Cl)ccc2o1  ...      0\n",
              "3  COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...  ...      0\n",
              "4  CC(=O)[C@H]1[C@H](C#N)C[C@H]2[C@@H]3CC=C4CC(O)...  ...      0\n",
              "\n",
              "[5 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F-i4wUBPnKFx"
      },
      "source": [
        "# Inference on the external dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XFdorY-KIp93",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 758
        },
        "outputId": "5c7b5f30-2ccc-4c4c-8a47-d68ff70bfe32"
      },
      "source": [
        "results_df = pl.predict_on_dataframe(test_data, contains_responses=True)\n",
        "results_df"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2021-08-11 14:46:24,270 Featurizing data...\n",
            "2021-08-11 14:46:24,276 Featurizing sample 0\n",
            "2021-08-11 14:46:24,386 Done\n",
            "2021-08-11 14:46:24,397 Evaluating current model\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "number of features: 1024\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "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>base_rdkit_smiles</th>\n",
              "      <th>actual</th>\n",
              "      <th>pred</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>compound_name</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>CHEMBL198</th>\n",
              "      <td>CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL3039453</th>\n",
              "      <td>C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL472566</th>\n",
              "      <td>Nc1nc2cc(Cl)ccc2o1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL44657</th>\n",
              "      <td>COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL4063685</th>\n",
              "      <td>CC(=O)[C@H]1[C@H](C#N)C[C@H]2[C@@H]3CC=C4CC(O)...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL1200777</th>\n",
              "      <td>N[C@H](Cc1cc(I)c(Oc2cc(I)c(O)c(I)c2)c(I)c1)C(=...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL452231</th>\n",
              "      <td>COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL640</th>\n",
              "      <td>CCN(CC)CCNC(=O)c1ccc(N)cc1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL95804</th>\n",
              "      <td>CCN(CC)CCNC(=O)c1ccc(NS(C)(=O)=O)cc1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL1231530</th>\n",
              "      <td>O=C(O)CCNC(=O)c1ccccc1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL373742</th>\n",
              "      <td>N=C(N)NCCC[C@H](NC(=O)[C@@H]1CCCN1C(=O)[C@@H]1...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL1200820</th>\n",
              "      <td>CCN1CCN(C(=O)N[C@@H](C(=O)N[C@@H]2C(=O)N3[C@@H...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL15677</th>\n",
              "      <td>O=C(O)Cc1ccccc1Oc1ccc(Cl)cc1Cl</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL403</th>\n",
              "      <td>CC1(C)[C@H](C(=O)O)N2C(=O)C[C@H]2S1(=O)=O</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL184</th>\n",
              "      <td>Nc1nc(O)c2ncn(COCCO)c2n1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL438</th>\n",
              "      <td>Cc1ccnc(NS(=O)(=O)c2ccc(N)cc2)n1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL4088943</th>\n",
              "      <td>Cc1ccc(CCCN[C@H](C)CO)cc1C(=O)NCC12CC3CC(CC(C3...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL35505</th>\n",
              "      <td>NNc1nnc(NN)c2ccccc12</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>CHEMBL1201203</th>\n",
              "      <td>CN1[C@H]2CC[C@@H]1C[C@H](OC(c1ccccc1)c1ccccc1)C2</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               base_rdkit_smiles  actual  pred\n",
              "compound_name                                                                 \n",
              "CHEMBL198      CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...     0.0     0\n",
              "CHEMBL3039453  C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]...     1.0     0\n",
              "CHEMBL472566                                  Nc1nc2cc(Cl)ccc2o1     0.0     0\n",
              "CHEMBL44657    COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...     0.0     1\n",
              "CHEMBL4063685  CC(=O)[C@H]1[C@H](C#N)C[C@H]2[C@@H]3CC=C4CC(O)...     0.0     1\n",
              "CHEMBL1200777  N[C@H](Cc1cc(I)c(Oc2cc(I)c(O)c(I)c2)c(I)c1)C(=...     1.0     1\n",
              "CHEMBL452231   COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...     0.0     1\n",
              "CHEMBL640                             CCN(CC)CCNC(=O)c1ccc(N)cc1     0.0     0\n",
              "CHEMBL95804                 CCN(CC)CCNC(=O)c1ccc(NS(C)(=O)=O)cc1     0.0     0\n",
              "CHEMBL1231530                             O=C(O)CCNC(=O)c1ccccc1     0.0     0\n",
              "CHEMBL373742   N=C(N)NCCC[C@H](NC(=O)[C@@H]1CCCN1C(=O)[C@@H]1...     0.0     0\n",
              "CHEMBL1200820  CCN1CCN(C(=O)N[C@@H](C(=O)N[C@@H]2C(=O)N3[C@@H...     0.0     1\n",
              "CHEMBL15677                       O=C(O)Cc1ccccc1Oc1ccc(Cl)cc1Cl     1.0     0\n",
              "CHEMBL403              CC1(C)[C@H](C(=O)O)N2C(=O)C[C@H]2S1(=O)=O     0.0     0\n",
              "CHEMBL184                               Nc1nc(O)c2ncn(COCCO)c2n1     0.0     0\n",
              "CHEMBL438                       Cc1ccnc(NS(=O)(=O)c2ccc(N)cc2)n1     0.0     0\n",
              "CHEMBL4088943  Cc1ccc(CCCN[C@H](C)CO)cc1C(=O)NCC12CC3CC(CC(C3...     1.0     0\n",
              "CHEMBL35505                                 NNc1nnc(NN)c2ccccc12     0.0     0\n",
              "CHEMBL1201203   CN1[C@H]2CC[C@@H]1C[C@H](OC(c1ccccc1)c1ccccc1)C2     1.0     0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2Y7Bc7avABke"
      },
      "source": [
        "### New AMPL version function can be used instead of the previous cell command"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "x8T1AIW5U5j_",
        "outputId": "49ef4ca5-a93e-41dc-e035-c3f9b6faf341"
      },
      "source": [
        "import os\n",
        "files = os.listdir()\n",
        "for f in files: \n",
        "  if (f.endswith(\".tar.gz\")):\n",
        "    mfile = f \n",
        "\n",
        "print(mfile)"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "morgan_warner_combined_bsep_data_model_c6623e4c-b43c-404d-9bd7-59b86fa138bc.tar.gz\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QOcw6dOEVhIH"
      },
      "source": [
        "## Replace the mfile with the file that is shown above. Note the filename will be different for each run "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 692
        },
        "id": "0INOGdDl__Yq",
        "outputId": "471fe19d-3ba1-4e3d-fbfd-daaf4f01b935"
      },
      "source": [
        "from atomsci.ddm.pipeline import predict_from_model as pfm\n",
        "# mfile = \"morgan_warner_combined_bsep_data_model_c6623e4c-b43c-404d-9bd7-59b86fa138bc.tar.gz\"\n",
        "mpath = '/content'\n",
        "input_df = test_data\n",
        "response_col = \"active\"\n",
        "smiles_col = \"base_rdkit_smiles\"\n",
        "pfm.predict_from_model_file(model_path = mfile, \n",
        "                            input_df = input_df, \n",
        "                            smiles_col = smiles_col, \n",
        "                            response_col = response_col)"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Standardizing SMILES strings for 19 compounds.\n",
            "Featurization = DynamicFeaturization with ecfp features\n",
            "number of features: 1024\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>compound_id</th>\n",
              "      <th>standardized_smiles</th>\n",
              "      <th>active_actual</th>\n",
              "      <th>active_prob</th>\n",
              "      <th>active_pred</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>compound_000000</td>\n",
              "      <td>CN(C)[C@@H]1C(O)=C(C(N)=O)C(=O)[C@@]2(O)C(O)=C...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.317333</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>compound_000001</td>\n",
              "      <td>C[C@]12C[C@H](O)[C@@]3(F)C(CCC4=CC(=O)C=C[C@@]...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.262000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>compound_000002</td>\n",
              "      <td>Nc1nc2cc(Cl)ccc2o1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.296000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>compound_000003</td>\n",
              "      <td>COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.513417</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>compound_000004</td>\n",
              "      <td>CC(=O)[C@H]1[C@H](C#N)C[C@H]2[C@@H]3CC=C4CC(O)...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.743000</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>compound_000005</td>\n",
              "      <td>N[C@H](Cc1cc(I)c(Oc2cc(I)c(O)c(I)c2)c(I)c1)C(=...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.800000</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>compound_000006</td>\n",
              "      <td>COc1cc([C@@H]2c3cc4c(cc3[C@@H](O[C@@H]3O[C@@H]...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.522000</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>compound_000007</td>\n",
              "      <td>CCN(CC)CCNC(=O)c1ccc(N)cc1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.134000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>compound_000008</td>\n",
              "      <td>CCN(CC)CCNC(=O)c1ccc(NS(C)(=O)=O)cc1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.260000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>compound_000009</td>\n",
              "      <td>O=C(O)CCNC(=O)c1ccccc1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.118000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>compound_000010</td>\n",
              "      <td>N=C(N)NCCC[C@H](NC(=O)[C@@H]1CCCN1C(=O)[C@@H]1...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.332000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>compound_000011</td>\n",
              "      <td>CCN1CCN(C(=O)N[C@@H](C(=O)N[C@@H]2C(=O)N3[C@@H...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.557667</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>compound_000012</td>\n",
              "      <td>O=C(O)Cc1ccccc1Oc1ccc(Cl)cc1Cl</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.420000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>compound_000013</td>\n",
              "      <td>CC1(C)[C@H](C(=O)O)N2C(=O)C[C@H]2S1(=O)=O</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.427000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>compound_000014</td>\n",
              "      <td>Nc1nc(O)c2ncn(COCCO)c2n1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.364000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>compound_000015</td>\n",
              "      <td>Cc1ccnc(NS(=O)(=O)c2ccc(N)cc2)n1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.341000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>compound_000016</td>\n",
              "      <td>Cc1ccc(CCCN[C@H](C)CO)cc1C(=O)NCC12CC3CC(CC(C3...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.233667</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>compound_000017</td>\n",
              "      <td>NNc1nnc(NN)c2ccccc12</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.192000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>compound_000018</td>\n",
              "      <td>CN1[C@H]2CC[C@@H]1C[C@H](OC(c1ccccc1)c1ccccc1)C2</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.152000</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        compound_id  ... active_pred\n",
              "0   compound_000000  ...           0\n",
              "1   compound_000001  ...           0\n",
              "2   compound_000002  ...           0\n",
              "3   compound_000003  ...           1\n",
              "4   compound_000004  ...           1\n",
              "5   compound_000005  ...           1\n",
              "6   compound_000006  ...           1\n",
              "7   compound_000007  ...           0\n",
              "8   compound_000008  ...           0\n",
              "9   compound_000009  ...           0\n",
              "10  compound_000010  ...           0\n",
              "11  compound_000011  ...           1\n",
              "12  compound_000012  ...           0\n",
              "13  compound_000013  ...           0\n",
              "14  compound_000014  ...           0\n",
              "15  compound_000015  ...           0\n",
              "16  compound_000016  ...           0\n",
              "17  compound_000017  ...           0\n",
              "18  compound_000018  ...           0\n",
              "\n",
              "[19 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W1SipNLBnP1p"
      },
      "source": [
        "# Report some performance metrics"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Iss-mPZuKoXx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0f8b3ff3-73c3-424e-8ade-da7c4318ef15"
      },
      "source": [
        "from sklearn.metrics import balanced_accuracy_score\n",
        "y_true = results_df['actual'].values\n",
        "y_pred = results_df['pred'].values\n",
        "balanced_accuracy_score(y_true, y_pred)"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.4571428571428572"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4cTVESd_6mtW",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "222735c0-a2a9-4db3-d2c8-aa1a6b39ef27"
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "confusion_matrix(y_true, y_pred)"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[10,  4],\n",
              "       [ 4,  1]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ISuW9k5-8pvs",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0fa1e242-a94f-4f22-8cb5-508f5adcf24d"
      },
      "source": [
        "!date"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Wed Aug 11 15:05:46 UTC 2021\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4PRAlkJT-3Gt"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}