VAE_MLP.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "machine_shape": "hm",
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "VAE-MLP\n",
        "\n",
        "This notebook contains the code that we tried to implement for the VAE-MLP model. The model is a combination of the earlier studies reported for MolGAN and Chemical VAE- we utilize the same model workflow as Chemical VAE and use an additional component by using graphical representation instead of SMILES and extract the adjacency matrix and a feature matrix which serve as the input to the encoder. The encoder has graph convolutional layers and forces the data through an informational bottleneck, the latent representation, and a decoder reconstructs the adjacency matrix and the feature matrix solely from this latent embedding. Additionally, an MLP model is trained to map between the latent representation to the desired property prediction. Training of all the 3 components of the model, the encoder, decoder and MLP takes place simulataneously on a particular dataset.\n",
        "\n",
        "Code source\n",
        "\n",
        "Initially we tried to implement the simple model that was cited in the ChemVAE paper, but faced some troubles incorporating all the required libraries owning to the fact that most of the code used older versions of libraries of Keras and Tensorflow. Most of the code is borrowed from another implementation of the code with the addition of the graphical representation from a demo document on Keras.io : https://keras.io/examples/generative/molecule_generation/ which uses a newer version of the libraries.\n",
        "\n",
        "This initial part is building the model and training the model on the Zinc database and is borrowed from the source cited earlier. We then tried to test the model property predictor by comparing the predicted properties with the ground truth on the training set."
      ],
      "metadata": {
        "id": "3EzwaRbzKiAm"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JD7NYNozKhUR",
        "outputId": "759b86bf-21ff-4982-cd09-ab22cf2081bb"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: PyTDC in /usr/local/lib/python3.10/dist-packages (0.4.1)\n",
            "Requirement already satisfied: rdkit-pypi in /usr/local/lib/python3.10/dist-packages (from PyTDC) (2021.9.4)\n",
            "Requirement already satisfied: fuzzywuzzy in /usr/local/lib/python3.10/dist-packages (from PyTDC) (0.18.0)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from PyTDC) (1.23.5)\n",
            "Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (from PyTDC) (1.5.3)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from PyTDC) (4.66.1)\n",
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.10/dist-packages (from PyTDC) (1.2.2)\n",
            "Requirement already satisfied: seaborn in /usr/local/lib/python3.10/dist-packages (from PyTDC) (0.12.2)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from PyTDC) (2.31.0)\n",
            "Requirement already satisfied: huggingface-hub in /usr/local/lib/python3.10/dist-packages (from PyTDC) (0.19.4)\n",
            "Requirement already satisfied: dataclasses in /usr/local/lib/python3.10/dist-packages (from PyTDC) (0.6)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from huggingface-hub->PyTDC) (3.13.1)\n",
            "Requirement already satisfied: fsspec>=2023.5.0 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub->PyTDC) (2023.6.0)\n",
            "Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub->PyTDC) (6.0.1)\n",
            "Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub->PyTDC) (4.5.0)\n",
            "Requirement already satisfied: packaging>=20.9 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub->PyTDC) (23.2)\n",
            "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.10/dist-packages (from pandas->PyTDC) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas->PyTDC) (2023.3.post1)\n",
            "Requirement already satisfied: Pillow in /usr/local/lib/python3.10/dist-packages (from rdkit-pypi->PyTDC) (9.4.0)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->PyTDC) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->PyTDC) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->PyTDC) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->PyTDC) (2023.11.17)\n",
            "Requirement already satisfied: scipy>=1.3.2 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->PyTDC) (1.8.1)\n",
            "Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->PyTDC) (1.3.2)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->PyTDC) (3.2.0)\n",
            "Requirement already satisfied: matplotlib!=3.6.1,>=3.1 in /usr/local/lib/python3.10/dist-packages (from seaborn->PyTDC) (3.7.1)\n",
            "Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.1->seaborn->PyTDC) (1.2.0)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.1->seaborn->PyTDC) (0.12.1)\n",
            "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.1->seaborn->PyTDC) (4.45.1)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.1->seaborn->PyTDC) (1.4.5)\n",
            "Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.1->seaborn->PyTDC) (3.1.1)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.1->pandas->PyTDC) (1.16.0)\n",
            "Requirement already satisfied: deepchem in /usr/local/lib/python3.10/dist-packages (2.7.1)\n",
            "Requirement already satisfied: joblib in /usr/local/lib/python3.10/dist-packages (from deepchem) (1.3.2)\n",
            "Requirement already satisfied: numpy>=1.21 in /usr/local/lib/python3.10/dist-packages (from deepchem) (1.23.5)\n",
            "Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (from deepchem) (1.5.3)\n",
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.10/dist-packages (from deepchem) (1.2.2)\n",
            "Requirement already satisfied: scipy<1.9 in /usr/local/lib/python3.10/dist-packages (from deepchem) (1.8.1)\n",
            "Requirement already satisfied: rdkit in /usr/local/lib/python3.10/dist-packages (from deepchem) (2023.9.2)\n",
            "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.10/dist-packages (from pandas->deepchem) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas->deepchem) (2023.3.post1)\n",
            "Requirement already satisfied: Pillow in /usr/local/lib/python3.10/dist-packages (from rdkit->deepchem) (9.4.0)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->deepchem) (3.2.0)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.1->pandas->deepchem) (1.16.0)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (1.8.1)\n",
            "Requirement already satisfied: numpy<1.25.0,>=1.17.3 in /usr/local/lib/python3.10/dist-packages (from scipy) (1.23.5)\n"
          ]
        }
      ],
      "source": [
        "!pip -q install rdkit-pypi==2021.9.4\n",
        "!pip install PyTDC\n",
        "!pip install deepchem\n",
        "!pip install scipy"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# importing all the required libraries for building the model\n",
        "\n",
        "import ast\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "from rdkit import Chem, RDLogger\n",
        "from rdkit.Chem import BondType\n",
        "from rdkit.Chem.Draw import MolsToGridImage\n",
        "from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score, median_absolute_error\n",
        "\n",
        "RDLogger.DisableLog(\"rdApp.*\")"
      ],
      "metadata": {
        "id": "LsAigh1UKlF9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Dataset\n",
        "\n",
        "We use the [**ZINC – A Free Database of Commercially Available Compounds for\n",
        "Virtual Screening**](https://bit.ly/3IVBI4x) dataset. The dataset comes with molecule\n",
        "formula in SMILE representation along with their respective molecular properties such as\n",
        "**logP** (water–octanal partition coefficient), **SAS** (synthetic\n",
        "accessibility score) and **QED** (Qualitative Estimate of Drug-likeness)."
      ],
      "metadata": {
        "id": "hyqYVOz6MQln"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "csv_path = keras.utils.get_file(\n",
        "    \"/content/250k_rndm_zinc_drugs_clean_3.csv\",\n",
        "    \"https://raw.githubusercontent.com/aspuru-guzik-group/chemical_vae/master/models/zinc_properties/250k_rndm_zinc_drugs_clean_3.csv\",\n",
        ")\n",
        "\n",
        "df = pd.read_csv(\"/content/250k_rndm_zinc_drugs_clean_3.csv\")\n",
        "df[\"smiles\"] = df[\"smiles\"].apply(lambda s: s.replace(\"\\n\", \"\"))\n",
        "df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "dIDTe4QxKmX_",
        "outputId": "0b3979ae-3077-455d-9d91-7c72ee13ecc0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                              smiles     logP       qed  \\\n",
              "0            CC(C)(C)c1ccc2occ(CC(=O)Nc3ccccc3F)c2c1  5.05060  0.702012   \n",
              "1       C[C@@H]1CC(Nc2cncc(-c3nncn3C)c2)C[C@@H](C)C1  3.11370  0.928975   \n",
              "2  N#Cc1ccc(-c2ccc(O[C@@H](C(=O)N3CCCC3)c3ccccc3)...  4.96778  0.599682   \n",
              "3  CCOC(=O)[C@@H]1CCCN(C(=O)c2nc(-c3ccc(C)cc3)n3c...  4.00022  0.690944   \n",
              "4  N#CC1=C(SCC(=O)Nc2cccc(Cl)c2)N=C([O-])[C@H](C#...  3.60956  0.789027   \n",
              "\n",
              "        SAS  \n",
              "0  2.084095  \n",
              "1  3.432004  \n",
              "2  2.470633  \n",
              "3  2.822753  \n",
              "4  4.035182  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-20e3e4d1-c412-49d8-8816-51cb0d59a33b\" class=\"colab-df-container\">\n",
              "    <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>smiles</th>\n",
              "      <th>logP</th>\n",
              "      <th>qed</th>\n",
              "      <th>SAS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>CC(C)(C)c1ccc2occ(CC(=O)Nc3ccccc3F)c2c1</td>\n",
              "      <td>5.05060</td>\n",
              "      <td>0.702012</td>\n",
              "      <td>2.084095</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>C[C@@H]1CC(Nc2cncc(-c3nncn3C)c2)C[C@@H](C)C1</td>\n",
              "      <td>3.11370</td>\n",
              "      <td>0.928975</td>\n",
              "      <td>3.432004</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>N#Cc1ccc(-c2ccc(O[C@@H](C(=O)N3CCCC3)c3ccccc3)...</td>\n",
              "      <td>4.96778</td>\n",
              "      <td>0.599682</td>\n",
              "      <td>2.470633</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>CCOC(=O)[C@@H]1CCCN(C(=O)c2nc(-c3ccc(C)cc3)n3c...</td>\n",
              "      <td>4.00022</td>\n",
              "      <td>0.690944</td>\n",
              "      <td>2.822753</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>N#CC1=C(SCC(=O)Nc2cccc(Cl)c2)N=C([O-])[C@H](C#...</td>\n",
              "      <td>3.60956</td>\n",
              "      <td>0.789027</td>\n",
              "      <td>4.035182</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-20e3e4d1-c412-49d8-8816-51cb0d59a33b')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-20e3e4d1-c412-49d8-8816-51cb0d59a33b button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-20e3e4d1-c412-49d8-8816-51cb0d59a33b');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-ffdfceda-41fb-4cf3-996a-b7c702324912\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-ffdfceda-41fb-4cf3-996a-b7c702324912')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-ffdfceda-41fb-4cf3-996a-b7c702324912 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oiR-eTB9KoA0",
        "outputId": "f593c552-19b5-4a29-dc4a-d74c37691913"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(249455, 4)"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "SMILE_CHARSET = '[\"C\", \"B\", \"F\", \"I\", \"H\", \"O\", \"N\", \"S\", \"P\", \"Cl\", \"Br\",\"Na\",\"Si\"]'    # Note here we added Na and Si to accomodate molecules in our external test set\n",
        "\n",
        "bond_mapping = {\"SINGLE\": 0, \"DOUBLE\": 1, \"TRIPLE\": 2, \"AROMATIC\": 3}\n",
        "bond_mapping.update(\n",
        "    {0: BondType.SINGLE, 1: BondType.DOUBLE, 2: BondType.TRIPLE, 3: BondType.AROMATIC}\n",
        ")\n",
        "SMILE_CHARSET = ast.literal_eval(SMILE_CHARSET)\n",
        "\n",
        "MAX_MOLSIZE = max(df[\"smiles\"].str.len())\n",
        "SMILE_to_index = dict((c, i) for i, c in enumerate(SMILE_CHARSET))\n",
        "index_to_SMILE = dict((i, c) for i, c in enumerate(SMILE_CHARSET))\n",
        "atom_mapping = dict(SMILE_to_index)\n",
        "atom_mapping.update(index_to_SMILE)\n",
        "\n",
        "BATCH_SIZE = 100\n",
        "EPOCHS = 10\n",
        "\n",
        "VAE_LR = 5e-4\n",
        "NUM_ATOMS = 120  # Maximum number of atoms\n",
        "\n",
        "ATOM_DIM = len(SMILE_CHARSET)  # Number of atom types\n",
        "BOND_DIM = 4 + 1  # Number of bond types\n",
        "LATENT_DIM = 435  # Size of the latent space\n",
        "\n",
        "\n",
        "\n",
        "# generate adjaceny and features matrix from the smiles structure\n",
        "\n",
        "def smiles_to_graph(smiles):\n",
        "    # Converts SMILES to molecule object\n",
        "    molecule = Chem.MolFromSmiles(smiles)\n",
        "\n",
        "    # Initialize adjacency and feature tensor\n",
        "    adjacency = np.zeros((BOND_DIM, NUM_ATOMS, NUM_ATOMS), \"float32\")\n",
        "    features = np.zeros((NUM_ATOMS, ATOM_DIM), \"float32\")\n",
        "\n",
        "    # loop over each atom in molecule\n",
        "    for atom in molecule.GetAtoms():\n",
        "        i = atom.GetIdx()\n",
        "        atom_type = atom_mapping[atom.GetSymbol()]\n",
        "        features[i] = np.eye(ATOM_DIM)[atom_type]\n",
        "        # loop over one-hop neighbors\n",
        "        for neighbor in atom.GetNeighbors():\n",
        "            j = neighbor.GetIdx()\n",
        "            bond = molecule.GetBondBetweenAtoms(i, j)\n",
        "            bond_type_idx = bond_mapping[bond.GetBondType().name]\n",
        "            adjacency[bond_type_idx, [i, j], [j, i]] = 1\n",
        "\n",
        "    # Where no bond, add 1 to last channel (indicating \"non-bond\")\n",
        "    # Notice: channels-first\n",
        "    adjacency[-1, np.sum(adjacency, axis=0) == 0] = 1\n",
        "\n",
        "    # Where no atom, add 1 to last column (indicating \"non-atom\")\n",
        "    features[np.where(np.sum(features, axis=1) == 0)[0], -1] = 1\n",
        "\n",
        "    return adjacency, features\n",
        "\n",
        "\n",
        "def graph_to_molecule(graph):\n",
        "    # Unpack graph\n",
        "    adjacency, features = graph\n",
        "\n",
        "    # RWMol is a molecule object intended to be edited\n",
        "    molecule = Chem.RWMol()\n",
        "\n",
        "    # Remove \"no atoms\" & atoms with no bonds\n",
        "    keep_idx = np.where(\n",
        "        (np.argmax(features, axis=1) != ATOM_DIM - 1)\n",
        "        & (np.sum(adjacency[:-1], axis=(0, 1)) != 0)\n",
        "    )[0]\n",
        "    features = features[keep_idx]\n",
        "    adjacency = adjacency[:, keep_idx, :][:, :, keep_idx]\n",
        "\n",
        "    # Add atoms to molecule\n",
        "    for atom_type_idx in np.argmax(features, axis=1):\n",
        "        atom = Chem.Atom(atom_mapping[atom_type_idx])\n",
        "        _ = molecule.AddAtom(atom)\n",
        "\n",
        "    # Add bonds between atoms in molecule; based on the upper triangles\n",
        "    # of the [symmetric] adjacency tensor\n",
        "    (bonds_ij, atoms_i, atoms_j) = np.where(np.triu(adjacency) == 1)\n",
        "    for (bond_ij, atom_i, atom_j) in zip(bonds_ij, atoms_i, atoms_j):\n",
        "        if atom_i == atom_j or bond_ij == BOND_DIM - 1:\n",
        "            continue\n",
        "        bond_type = bond_mapping[bond_ij]\n",
        "        molecule.AddBond(int(atom_i), int(atom_j), bond_type)\n",
        "\n",
        "    # Sanitize the molecule; for more information on sanitization, see\n",
        "    # https://www.rdkit.org/docs/RDKit_Book.html#molecular-sanitization\n",
        "    flag = Chem.SanitizeMol(molecule, catchErrors=True)\n",
        "    # Let's be strict. If sanitization fails, return None\n",
        "    if flag != Chem.SanitizeFlags.SANITIZE_NONE:\n",
        "        return None\n",
        "\n",
        "    return molecule"
      ],
      "metadata": {
        "id": "n7MTOD2gKqAF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Sample 75% of the data to create a training dataset. The random state ensures reproducibility.\n",
        "train_df = df.sample(frac=0.75, random_state=42)\n",
        "train_df.reset_index(drop=True, inplace=True)\n",
        "\n",
        "# Initialize lists to store the adjacency matrix, feature matrix, and QED (Quantitative Estimate of Drug-likeness) values.\n",
        "adjacency_tensor, feature_tensor, qed_tensor = [], [], []\n",
        "\n",
        "# Loop through the first 8000 samples in the training dataset.\n",
        "for idx in range(8000):\n",
        "    # Convert SMILES to graph representation: adjacency matrix and feature matrix.\n",
        "    adjacency, features = smiles_to_graph(train_df.loc[idx][\"smiles\"])\n",
        "    # Extract QED value of the molecule.\n",
        "    qed = train_df.loc[idx][\"qed\"]\n",
        "    # Append the extracted data to their respective lists.\n",
        "    adjacency_tensor.append(adjacency)\n",
        "    feature_tensor.append(features)\n",
        "    qed_tensor.append(qed)\n",
        "\n",
        "# Convert lists to numpy arrays for easier manipulation and to feed into the neural network.\n",
        "adjacency_tensor = np.array(adjacency_tensor)\n",
        "feature_tensor = np.array(feature_tensor)\n",
        "qed_tensor = np.array(qed_tensor)\n",
        "\n",
        "# Define a custom Keras layer for a Relational Graph Convolutional Network (R-GCN).\n",
        "class RelationalGraphConvLayer(keras.layers.Layer):\n",
        "    def __init__(\n",
        "        self,\n",
        "        units=128,  # Number of units in the dense layer.\n",
        "        activation=\"relu\",  # Activation function.\n",
        "        use_bias=False,  # Whether to use a bias vector.\n",
        "        kernel_initializer=\"glorot_uniform\",  # Initializer for the kernel weights.\n",
        "        bias_initializer=\"zeros\",  # Initializer for the bias vector.\n",
        "        kernel_regularizer=None,  # Regularizer function applied to the kernel weights matrix.\n",
        "        bias_regularizer=None,  # Regularizer function applied to the bias vector.\n",
        "        **kwargs\n",
        "    ):\n",
        "        super().__init__(**kwargs)\n",
        "        # Initialize layer attributes.\n",
        "        self.units = units\n",
        "        self.activation = keras.activations.get(activation)\n",
        "        self.use_bias = use_bias\n",
        "        self.kernel_initializer = keras.initializers.get(kernel_initializer)\n",
        "        self.bias_initializer = keras.initializers.get(bias_initializer)\n",
        "        self.kernel_regularizer = keras.regularizers.get(kernel_regularizer)\n",
        "        self.bias_regularizer = keras.regularizers.get(bias_regularizer)\n",
        "\n",
        "    def build(self, input_shape):\n",
        "        # Define the shapes of the weights and biases based on input shapes.\n",
        "        bond_dim = input_shape[0][1]\n",
        "        atom_dim = input_shape[1][2]\n",
        "\n",
        "        # Create a weight matrix for the layer.\n",
        "        self.kernel = self.add_weight(\n",
        "            shape=(bond_dim, atom_dim, self.units),\n",
        "            initializer=self.kernel_initializer,\n",
        "            regularizer=self.kernel_regularizer,\n",
        "            trainable=True,\n",
        "            name=\"W\",\n",
        "            dtype=tf.float32,\n",
        "        )\n",
        "\n",
        "        # Create a bias vector, if use_bias is True.\n",
        "        if self.use_bias:\n",
        "            self.bias = self.add_weight(\n",
        "                shape=(bond_dim, 1, self.units),\n",
        "                initializer=self.bias_initializer,\n",
        "                regularizer=self.bias_regularizer,\n",
        "                trainable=True,\n",
        "                name=\"b\",\n",
        "                dtype=tf.float32,\n",
        "            )\n",
        "\n",
        "        self.built = True\n",
        "\n",
        "    def call(self, inputs, training=False):\n",
        "        adjacency, features = inputs\n",
        "        # Perform graph convolution: aggregate information from neighbors.\n",
        "        x = tf.matmul(adjacency, features[:, None, :, :])\n",
        "        # Apply a linear transformation using the kernel.\n",
        "        x = tf.matmul(x, self.kernel)\n",
        "        if self.use_bias:\n",
        "            # Add the bias, if applicable.\n",
        "            x += self.bias\n",
        "        # Sum over the bond types dimension.\n",
        "        x_reduced = tf.reduce_sum(x, axis=1)\n",
        "        # Apply the activation function.\n",
        "        return self.activation(x_reduced)"
      ],
      "metadata": {
        "id": "_rIQESx1KryJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def get_encoder(\n",
        "    gconv_units, latent_dim, adjacency_shape, feature_shape, dense_units, dropout_rate\n",
        "):\n",
        "    adjacency = keras.layers.Input(shape=adjacency_shape)\n",
        "    features = keras.layers.Input(shape=feature_shape)\n",
        "\n",
        "    # Propagate through one or more graph convolutional layers\n",
        "    features_transformed = features\n",
        "    for units in gconv_units:\n",
        "        features_transformed = RelationalGraphConvLayer(units)(\n",
        "            [adjacency, features_transformed]\n",
        "        )\n",
        "    # Reduce 2-D representation of molecule to 1-D\n",
        "    x = keras.layers.GlobalAveragePooling1D()(features_transformed)\n",
        "\n",
        "    # Propagate through one or more densely connected layers\n",
        "    for units in dense_units:\n",
        "        x = layers.Dense(units, activation=\"relu\")(x)\n",
        "        x = layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    z_mean = layers.Dense(latent_dim, dtype=\"float32\", name=\"z_mean\")(x)\n",
        "    log_var = layers.Dense(latent_dim, dtype=\"float32\", name=\"log_var\")(x)\n",
        "\n",
        "    encoder = keras.Model([adjacency, features], [z_mean, log_var], name=\"encoder\")\n",
        "\n",
        "    return encoder\n",
        "\n",
        "\n",
        "def get_decoder(dense_units, dropout_rate, latent_dim, adjacency_shape, feature_shape):\n",
        "    latent_inputs = keras.Input(shape=(latent_dim,))\n",
        "\n",
        "    x = latent_inputs\n",
        "    for units in dense_units:\n",
        "        x = keras.layers.Dense(units, activation=\"tanh\")(x)\n",
        "        x = keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # Map outputs of previous layer (x) to [continuous] adjacency tensors (x_adjacency)\n",
        "    x_adjacency = keras.layers.Dense(tf.math.reduce_prod(adjacency_shape))(x)\n",
        "    x_adjacency = keras.layers.Reshape(adjacency_shape)(x_adjacency)\n",
        "    # Symmetrify tensors in the last two dimensions\n",
        "    x_adjacency = (x_adjacency + tf.transpose(x_adjacency, (0, 1, 3, 2))) / 2\n",
        "    x_adjacency = keras.layers.Softmax(axis=1)(x_adjacency)\n",
        "\n",
        "    # Map outputs of previous layer (x) to [continuous] feature tensors (x_features)\n",
        "    x_features = keras.layers.Dense(tf.math.reduce_prod(feature_shape))(x)\n",
        "    x_features = keras.layers.Reshape(feature_shape)(x_features)\n",
        "    x_features = keras.layers.Softmax(axis=2)(x_features)\n",
        "\n",
        "    decoder = keras.Model(\n",
        "        latent_inputs, outputs=[x_adjacency, x_features], name=\"decoder\"\n",
        "    )\n",
        "\n",
        "    return decoder"
      ],
      "metadata": {
        "id": "hlZqabulKtcI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# importance of reparemetrization trick\n",
        "class Sampling(layers.Layer):\n",
        "    def call(self, inputs):\n",
        "        z_mean, z_log_var = inputs\n",
        "        batch = tf.shape(z_log_var)[0]\n",
        "        dim = tf.shape(z_log_var)[1]\n",
        "        epsilon = tf.keras.backend.random_normal(shape=(batch, dim))\n",
        "        return z_mean + tf.exp(0.5 * z_log_var) * epsilon"
      ],
      "metadata": {
        "id": "rJpGjE3BKvQq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Defining the MoleculeGenerator class for training the model\n",
        "\n",
        "\n",
        "class MoleculeGenerator(keras.Model):\n",
        "    def __init__(self, encoder, decoder, max_len, **kwargs):\n",
        "        super().__init__(**kwargs)\n",
        "        self.encoder = encoder\n",
        "        self.decoder = decoder\n",
        "        self.property_prediction_layer = layers.Dense(1)\n",
        "        self.max_len = max_len\n",
        "\n",
        "        self.train_total_loss_tracker = keras.metrics.Mean(name=\"train_total_loss\")\n",
        "        self.val_total_loss_tracker = keras.metrics.Mean(name=\"val_total_loss\")\n",
        "\n",
        "    def train_step(self, data):\n",
        "        adjacency_tensor, feature_tensor, qed_tensor = data[0]\n",
        "        graph_real = [adjacency_tensor, feature_tensor]\n",
        "        self.batch_size = tf.shape(qed_tensor)[0]\n",
        "        with tf.GradientTape() as tape:\n",
        "            z_mean, z_log_var, qed_pred, gen_adjacency, gen_features = self(\n",
        "                graph_real, training=True\n",
        "            )\n",
        "            graph_generated = [gen_adjacency, gen_features]\n",
        "            total_loss = self._compute_loss(\n",
        "                z_log_var, z_mean, qed_tensor, qed_pred, graph_real, graph_generated\n",
        "            )\n",
        "\n",
        "        grads = tape.gradient(total_loss, self.trainable_weights)\n",
        "        self.optimizer.apply_gradients(zip(grads, self.trainable_weights))\n",
        "\n",
        "        self.train_total_loss_tracker.update_state(total_loss)\n",
        "        return {\"loss\": self.train_total_loss_tracker.result()}\n",
        "\n",
        "    def _compute_loss(\n",
        "        self, z_log_var, z_mean, qed_true, qed_pred, graph_real, graph_generated\n",
        "    ):\n",
        "\n",
        "        adjacency_real, features_real = graph_real\n",
        "        adjacency_gen, features_gen = graph_generated\n",
        "\n",
        "        adjacency_loss = tf.reduce_mean(\n",
        "            tf.reduce_sum(\n",
        "                keras.losses.categorical_crossentropy(adjacency_real, adjacency_gen),\n",
        "                axis=(1, 2),\n",
        "            )\n",
        "        )\n",
        "        features_loss = tf.reduce_mean(\n",
        "            tf.reduce_sum(\n",
        "                keras.losses.categorical_crossentropy(features_real, features_gen),\n",
        "                axis=(1),\n",
        "            )\n",
        "        )\n",
        "        kl_loss = -0.5 * tf.reduce_sum(\n",
        "            1 + z_log_var - tf.square(z_mean) - tf.exp(z_log_var), 1\n",
        "        )\n",
        "        kl_loss = tf.reduce_mean(kl_loss)\n",
        "\n",
        "        property_loss = tf.reduce_mean(\n",
        "            keras.losses.binary_crossentropy(qed_true, qed_pred)\n",
        "        )\n",
        "\n",
        "        graph_loss = self._gradient_penalty(graph_real, graph_generated)\n",
        "\n",
        "        # Note the loss is composed of 4 terms\n",
        "        # Minimizing the kl loss to restrict the latent space\n",
        "        # Plus the property loss to minimize the error of prediction of the desired property\n",
        "        # Plus the graph loss to minimize the gradient penalty\n",
        "        # Plus the adjacency loss and features loss to minimize the reconstruction loss from the input\n",
        "\n",
        "        return kl_loss + property_loss + graph_loss + adjacency_loss + features_loss\n",
        "\n",
        "    def _gradient_penalty(self, graph_real, graph_generated):\n",
        "        # Unpack graphs\n",
        "        adjacency_real, features_real = graph_real\n",
        "        adjacency_generated, features_generated = graph_generated\n",
        "\n",
        "        # Generate interpolated graphs (adjacency_interp and features_interp)\n",
        "        alpha = tf.random.uniform([self.batch_size])\n",
        "        alpha = tf.reshape(alpha, (self.batch_size, 1, 1, 1))\n",
        "        adjacency_interp = (adjacency_real * alpha) + (1 - alpha) * adjacency_generated\n",
        "        alpha = tf.reshape(alpha, (self.batch_size, 1, 1))\n",
        "        features_interp = (features_real * alpha) + (1 - alpha) * features_generated\n",
        "\n",
        "        # Compute the logits of interpolated graphs\n",
        "        with tf.GradientTape() as tape:\n",
        "            tape.watch(adjacency_interp)\n",
        "            tape.watch(features_interp)\n",
        "            _, _, logits, _, _ = self(\n",
        "                [adjacency_interp, features_interp], training=True\n",
        "            )\n",
        "\n",
        "        # Compute the gradients with respect to the interpolated graphs\n",
        "        grads = tape.gradient(logits, [adjacency_interp, features_interp])\n",
        "        # Compute the gradient penalty\n",
        "        grads_adjacency_penalty = (1 - tf.norm(grads[0], axis=1)) ** 2\n",
        "        grads_features_penalty = (1 - tf.norm(grads[1], axis=2)) ** 2\n",
        "        return tf.reduce_mean(\n",
        "            tf.reduce_mean(grads_adjacency_penalty, axis=(-2, -1))\n",
        "            + tf.reduce_mean(grads_features_penalty, axis=(-1))\n",
        "        )\n",
        "\n",
        "    def inference(self, batch_size):\n",
        "        z = tf.random.normal((batch_size, LATENT_DIM))\n",
        "        reconstruction_adjacency, reconstruction_features = model.decoder.predict(z)\n",
        "        # obtain one-hot encoded adjacency tensor\n",
        "        adjacency = tf.argmax(reconstruction_adjacency, axis=1)\n",
        "        adjacency = tf.one_hot(adjacency, depth=BOND_DIM, axis=1)\n",
        "        # Remove potential self-loops from adjacency\n",
        "        adjacency = tf.linalg.set_diag(adjacency, tf.zeros(tf.shape(adjacency)[:-1]))\n",
        "        # obtain one-hot encoded feature tensor\n",
        "        features = tf.argmax(reconstruction_features, axis=2)\n",
        "        features = tf.one_hot(features, depth=ATOM_DIM, axis=2)\n",
        "        return [\n",
        "            graph_to_molecule([adjacency[i].numpy(), features[i].numpy()])\n",
        "            for i in range(batch_size)\n",
        "        ]\n",
        "\n",
        "    def call(self, inputs):\n",
        "        z_mean, log_var = self.encoder(inputs)\n",
        "        z = Sampling()([z_mean, log_var])\n",
        "\n",
        "        gen_adjacency, gen_features = self.decoder(z)\n",
        "\n",
        "        property_pred = self.property_prediction_layer(z_mean)\n",
        "\n",
        "        return z_mean, log_var, property_pred, gen_adjacency, gen_features"
      ],
      "metadata": {
        "id": "qRJ1wgjWKyvC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# training the model on the Zinc database\n",
        "\n",
        "vae_optimizer = tf.keras.optimizers.Adam(learning_rate=VAE_LR)\n",
        "\n",
        "encoder = get_encoder(\n",
        "    gconv_units=[9],\n",
        "    adjacency_shape=(BOND_DIM, NUM_ATOMS, NUM_ATOMS),\n",
        "    feature_shape=(NUM_ATOMS, ATOM_DIM),\n",
        "    latent_dim=LATENT_DIM,\n",
        "    dense_units=[512],\n",
        "    dropout_rate=0.0,\n",
        ")\n",
        "decoder = get_decoder(\n",
        "    dense_units=[128, 256, 512],\n",
        "    dropout_rate=0.2,\n",
        "    latent_dim=LATENT_DIM,\n",
        "    adjacency_shape=(BOND_DIM, NUM_ATOMS, NUM_ATOMS),\n",
        "    feature_shape=(NUM_ATOMS, ATOM_DIM),\n",
        ")\n",
        "\n",
        "model = MoleculeGenerator(encoder, decoder, MAX_MOLSIZE)\n",
        "\n",
        "model.compile(vae_optimizer)\n",
        "history = model.fit([adjacency_tensor, feature_tensor, qed_tensor], epochs=EPOCHS)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XUmzMsrqKzUf",
        "outputId": "7eb2fbec-48a4-4431-992a-b4af95012c80"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "250/250 [==============================] - 16s 22ms/step - loss: 69106.4533\n",
            "Epoch 2/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 69030.3174\n",
            "Epoch 3/10\n",
            "250/250 [==============================] - 5s 22ms/step - loss: 69175.1844\n",
            "Epoch 4/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 69616.1142\n",
            "Epoch 5/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 69805.8854\n",
            "Epoch 6/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 69496.2684\n",
            "Epoch 7/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 69728.0669\n",
            "Epoch 8/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 70791.7982\n",
            "Epoch 9/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 72306.8904\n",
            "Epoch 10/10\n",
            "250/250 [==============================] - 6s 22ms/step - loss: 77130.5371\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# save the model\n",
        "NAME=\"ZINC250_Trained_v1\"\n",
        "\n",
        "model.save_weights(NAME)"
      ],
      "metadata": {
        "id": "V5cuuGNaK1ZD"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# see the R^2 value for this task with the Zinc dataset\n",
        "\n",
        "y_pred = model.predict([adjacency_tensor[:500], feature_tensor[:500]])[2]\n",
        "\n",
        "r2 = r2_score(qed_tensor[:500],y_pred)\n",
        "print(\"R-squared value:\", r2)"
      ],
      "metadata": {
        "id": "bcZqBN6bK4ib",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7230ba0e-803f-4a47-cc4c-73eaacf77e07"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "16/16 [==============================] - 0s 6ms/step\n",
            "R-squared value: -9094.775578850782\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Judging by the performance of the property predictor of QED on the Zinc database in terms of the R^2 value on the training set, it seems that even the base model trained on a big database of 250k molecules needs a lot of tuning, which we were not able to do so in the scope of this project."
      ],
      "metadata": {
        "id": "jehu8b1oK9Mv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# accessing the csv training and external test set data from google drive\n",
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive',force_remount=True)"
      ],
      "metadata": {
        "id": "iMZJOW2UK-tF",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9acdc714-09f1-4d77-adeb-1bb09404d839"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/gdrive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.manifold import TSNE\n",
        "from tdc.single_pred import ADME\n",
        "from rdkit import Chem\n",
        "from rdkit.Chem import Descriptors\n",
        "from rdkit.ML.Descriptors import MoleculeDescriptors\n",
        "\n",
        "# Load datasets\n",
        "filepath='/content/gdrive/'\n",
        "\n",
        "\n",
        "data = ADME(name = 'Caco2_Wang')\n",
        "df_new = data.get_data()"
      ],
      "metadata": {
        "id": "wwX88ieiLR6c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "331b6294-b6d5-4c64-f766-9101cd359ffd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Downloading...\n",
            "100%|██████████| 82.5k/82.5k [00:00<00:00, 4.00MiB/s]\n",
            "Loading...\n",
            "Done!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_new.head()"
      ],
      "metadata": {
        "id": "YXnCbMr0LXC2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "3384bf49-a67f-49ad-faf5-41db79196c1f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                             Drug_ID  \\\n",
              "0                                    (-)-epicatechin   \n",
              "1  (2E,4Z,8Z)-N-isobutyldodeca-2,4,10-triene-8 -y...   \n",
              "2                                            codeine   \n",
              "3                                         creatinine   \n",
              "4                                            danazol   \n",
              "\n",
              "                                                Drug         Y  \n",
              "0            Oc1cc(O)c2c(c1)OC(c1ccc(O)c(O)c1)C(O)C2 -6.220000  \n",
              "1                   C/C=C\\C#CCC/C=C\\C=C\\C(=O)NCC(C)C -3.860000  \n",
              "2  COc1ccc2c3c1O[C@H]1[C@@H](O)C=C[C@H]4[C@@H](C2... -4.090000  \n",
              "3                                     CN1CC(=O)NC1=N -5.935409  \n",
              "4  C#C[C@]1(O)CC[C@H]2[C@@H]3CCC4=Cc5oncc5C[C@]4(... -4.840000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-4ed4cd1b-cd1c-4c0f-962c-5c1ccb9335fd\" class=\"colab-df-container\">\n",
              "    <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>Drug_ID</th>\n",
              "      <th>Drug</th>\n",
              "      <th>Y</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>(-)-epicatechin</td>\n",
              "      <td>Oc1cc(O)c2c(c1)OC(c1ccc(O)c(O)c1)C(O)C2</td>\n",
              "      <td>-6.220000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>(2E,4Z,8Z)-N-isobutyldodeca-2,4,10-triene-8 -y...</td>\n",
              "      <td>C/C=C\\C#CCC/C=C\\C=C\\C(=O)NCC(C)C</td>\n",
              "      <td>-3.860000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>codeine</td>\n",
              "      <td>COc1ccc2c3c1O[C@H]1[C@@H](O)C=C[C@H]4[C@@H](C2...</td>\n",
              "      <td>-4.090000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>creatinine</td>\n",
              "      <td>CN1CC(=O)NC1=N</td>\n",
              "      <td>-5.935409</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>danazol</td>\n",
              "      <td>C#C[C@]1(O)CC[C@H]2[C@@H]3CCC4=Cc5oncc5C[C@]4(...</td>\n",
              "      <td>-4.840000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-4ed4cd1b-cd1c-4c0f-962c-5c1ccb9335fd')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-4ed4cd1b-cd1c-4c0f-962c-5c1ccb9335fd button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-4ed4cd1b-cd1c-4c0f-962c-5c1ccb9335fd');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-63fa9fc2-40cb-49f2-83ab-b291b5a4c175\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-63fa9fc2-40cb-49f2-83ab-b291b5a4c175')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-63fa9fc2-40cb-49f2-83ab-b291b5a4c175 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# defining the adjacency matrix, feature matrix and the Y_tensor for the property predictor\n",
        "\n",
        "train_df_new = df_new.sample(frac=0.75, random_state=42)  # random state is a seed value\n",
        "\n",
        "len_train_df_new=train_df_new.shape[0]\n",
        "train_df_new.reset_index(drop=True, inplace=True)\n",
        "\n",
        "adjacency_tensor_new, feature_tensor_new, Y_tensor = [], [], []\n",
        "\n",
        "\n",
        "for idx in range(len_train_df_new):\n",
        "    adjacency, features = smiles_to_graph(train_df_new.loc[idx][\"Drug\"])\n",
        "    Y= train_df_new.loc[idx][\"Y\"]\n",
        "    adjacency_tensor_new.append(adjacency)\n",
        "    feature_tensor_new.append(features)\n",
        "    Y_tensor.append(Y)\n",
        "\n",
        "adjacency_tensor_new = np.array(adjacency_tensor_new)\n",
        "feature_tensor_new= np.array(feature_tensor_new)\n",
        "Y_tensor = np.array(Y_tensor)"
      ],
      "metadata": {
        "id": "ahJjTEQSLemR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(f\" Min of qed is {np.min(qed_tensor)}, max of qed is {np.max(qed_tensor)}, min of Y is {np.min(Y_tensor)}, max of Y is {np.max(Y_tensor)} \")"
      ],
      "metadata": {
        "id": "WT3bt2_TLh0G",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1f9386fb-7076-4556-c8a1-03c92ad8a8b7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " Min of qed is 0.155273197945, max of qed is 0.946694656878, min of Y is -7.7600002, max of Y is -3.51 \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(Y_tensor)"
      ],
      "metadata": {
        "id": "DxoqoooJLjlU",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 517
        },
        "outputId": "0b3bb272-ffaf-4d56-af2e-e87d73d30c7c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array([ 11.,  21.,  26.,  52., 102., 113., 134., 163.,  54.,   6.]),\n",
              " array([-7.7600002 , -7.33500018, -6.91000016, -6.48500014, -6.06000012,\n",
              "        -5.6350001 , -5.21000008, -4.78500006, -4.36000004, -3.93500002,\n",
              "        -3.51      ]),\n",
              " <BarContainer object of 10 artists>)"
            ]
          },
          "metadata": {},
          "execution_count": 18
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(qed_tensor)"
      ],
      "metadata": {
        "id": "s6hKkA10LkL3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 534
        },
        "outputId": "a3aeb36f-519e-4d4d-c864-742405014da2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array([  11.,   49.,  156.,  282.,  473.,  791., 1195., 1731., 2185.,\n",
              "        1127.]),\n",
              " array([0.1552732 , 0.23441534, 0.31355749, 0.39269964, 0.47184178,\n",
              "        0.55098393, 0.63012607, 0.70926822, 0.78841037, 0.86755251,\n",
              "        0.94669466]),\n",
              " <BarContainer object of 10 artists>)"
            ]
          },
          "metadata": {},
          "execution_count": 19
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import MinMaxScaler\n",
        "\n",
        "scaler = MinMaxScaler(feature_range=(np.min(qed_tensor), np.max(qed_tensor)))\n",
        "\n",
        "Y_tensor_rescaled=scaler.fit_transform(Y_tensor.reshape(-1, 1))\n",
        "\n",
        "print(Y_tensor_rescaled)"
      ],
      "metadata": {
        "id": "Ubu5wjf_Llj5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Initialize a new instance of MoleculeGenerator\n",
        "new_model = MoleculeGenerator(encoder, decoder, MAX_MOLSIZE)\n",
        "new_model.property_prediction_layer = keras.layers.Dense(1, activation='linear')\n",
        "\n",
        "# Load weights into the new model, excluding the predictor layer\n",
        "new_model.load_weights(f'{NAME}')\n",
        "\n",
        "# Freeze the encoder and decoder layers\n",
        "for layer in new_model.encoder.layers:\n",
        "    layer.trainable = False\n",
        "for layer in new_model.decoder.layers:\n",
        "    layer.trainable = False"
      ],
      "metadata": {
        "id": "Mw8lT3SZLnCB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# new_model.compile(vae_optimizer)\n",
        "# Compile the new model, that is trained on the Zinc database\n",
        "new_model.compile(optimizer='adam', loss='mean_squared_error')\n",
        "\n",
        "adjacency_tensor = adjacency_tensor[:682]\n",
        "feature_tensor = feature_tensor[:682]\n",
        "\n",
        "# new_model.compile(vae_optimizer)\n",
        "history_2 = new_model.fit([adjacency_tensor, feature_tensor, Y_tensor_rescaled], epochs=EPOCHS)"
      ],
      "metadata": {
        "id": "Z3GrIpUQLpHh",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "aa0aa24b-9e0e-44b2-babb-39b64e5741df"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "22/22 [==============================] - 2s 14ms/step - loss: 80120.2802\n",
            "Epoch 2/10\n",
            "22/22 [==============================] - 0s 12ms/step - loss: 80161.5802\n",
            "Epoch 3/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80180.2160\n",
            "Epoch 4/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80120.3533\n",
            "Epoch 5/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80140.5265\n",
            "Epoch 6/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80077.3485\n",
            "Epoch 7/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80119.6613\n",
            "Epoch 8/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80107.1189\n",
            "Epoch 9/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80083.6726\n",
            "Epoch 10/10\n",
            "22/22 [==============================] - 0s 11ms/step - loss: 80155.5245\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# R^2 value on the training set\n",
        "\n",
        "\n",
        "y_pred = new_model.predict([adjacency_tensor_new, feature_tensor_new])[2]\n",
        "\n",
        "r2 = r2_score(Y_tensor, y_pred)\n",
        "print(\"R-squared value:\", r2)"
      ],
      "metadata": {
        "id": "MQlAl4JrLsFj",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c418561e-e2ae-4e76-a686-f17d431cb676"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "22/22 [==============================] - 0s 8ms/step\n",
            "R-squared value: -469.85410100405335\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.scatter(Y_tensor_rescaled,y_pred,color='blue', edgecolor='black')\n",
        "plt.xlabel('Actual values rescaled')\n",
        "plt.ylabel('Predicted values')\n",
        "\n",
        "x = np.linspace(plt.xlim()[0], plt.xlim()[1], 100)\n",
        "plt.plot(x, x, label='y=x line', color='black',linestyle='dashed')"
      ],
      "metadata": {
        "id": "cVmnaLd9Ls8l",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "outputId": "97174d43-c647-4ff7-85df-c82822fd924f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7c8166fdbca0>]"
            ]
          },
          "metadata": {},
          "execution_count": 48
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Now predicting properties without transfer learning\n",
        "\n",
        "vae_optimizer = tf.keras.optimizers.Adam(learning_rate=VAE_LR)\n",
        "\n",
        "encoder = get_encoder(\n",
        "    gconv_units=[9],\n",
        "    adjacency_shape=(BOND_DIM, NUM_ATOMS, NUM_ATOMS),\n",
        "    feature_shape=(NUM_ATOMS, ATOM_DIM),\n",
        "    latent_dim=LATENT_DIM,\n",
        "    dense_units=[512],\n",
        "    dropout_rate=0.0,\n",
        ")\n",
        "decoder = get_decoder(\n",
        "    dense_units=[128, 256, 512],\n",
        "    dropout_rate=0.2,\n",
        "    latent_dim=LATENT_DIM,\n",
        "    adjacency_shape=(BOND_DIM, NUM_ATOMS, NUM_ATOMS),\n",
        "    feature_shape=(NUM_ATOMS, ATOM_DIM),\n",
        ")\n",
        "\n",
        "model = MoleculeGenerator(encoder, decoder, MAX_MOLSIZE)\n",
        "\n",
        "model.compile(vae_optimizer)\n",
        "history = model.fit([adjacency_tensor_new, feature_tensor_new, Y_tensor_rescaled], epochs=EPOCHS)"
      ],
      "metadata": {
        "id": "ovZKrd_0Lv8z",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1cdafc28-a68e-4fcf-b8e3-37ceefb9bfcd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "22/22 [==============================] - 4s 25ms/step - loss: 69329.2874\n",
            "Epoch 2/10\n",
            "22/22 [==============================] - 1s 23ms/step - loss: 69302.8349\n",
            "Epoch 3/10\n",
            "22/22 [==============================] - 1s 23ms/step - loss: 69004.4229\n",
            "Epoch 4/10\n",
            "22/22 [==============================] - 1s 24ms/step - loss: 68931.4375\n",
            "Epoch 5/10\n",
            "22/22 [==============================] - 1s 24ms/step - loss: 68932.0605\n",
            "Epoch 6/10\n",
            "22/22 [==============================] - 1s 24ms/step - loss: 68933.6362\n",
            "Epoch 7/10\n",
            "22/22 [==============================] - 1s 23ms/step - loss: 68938.7503\n",
            "Epoch 8/10\n",
            "22/22 [==============================] - 1s 24ms/step - loss: 68951.9810\n",
            "Epoch 9/10\n",
            "22/22 [==============================] - 1s 23ms/step - loss: 68955.9803\n",
            "Epoch 10/10\n",
            "22/22 [==============================] - 1s 24ms/step - loss: 68971.0041\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred_noT = model.predict([adjacency_tensor_new, feature_tensor_new])[2]\n",
        "\n",
        "from sklearn.metrics import r2_score\n",
        "\n",
        "r2 = r2_score(Y_tensor_rescaled, y_pred_noT)\n",
        "print(\"R-squared value:\", r2)"
      ],
      "metadata": {
        "id": "-vg8hKxcLwvD",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "72d7bf84-5171-45f4-fd5f-f4e1401044e0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "22/22 [==============================] - 0s 8ms/step\n",
            "R-squared value: -0.6714782606871863\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.scatter(Y_tensor_rescaled,y_pred_noT,color='blue', edgecolor='black')\n",
        "plt.xlabel('Actual values rescaled')\n",
        "plt.ylabel('Predicted values')\n",
        "\n",
        "x = np.linspace(plt.xlim()[0], plt.xlim()[1], 100)\n",
        "plt.plot(x, x, label='y=x line', color='black',linestyle='dashed')"
      ],
      "metadata": {
        "id": "ntFKUguPLyz5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "outputId": "fa7ed57f-8f49-472c-880b-3a15cd610dfd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7c8167302800>]"
            ]
          },
          "metadata": {},
          "execution_count": 49
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Now checking the performance on the external test set\n",
        "\n",
        "df_test = pd.read_csv('external.csv')\n",
        "df_test.head()"
      ],
      "metadata": {
        "id": "9OC6XACjL0gd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "13b3d879-8ce0-47bc-ab86-cf9e003a42d7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     Molecule ID                                             SMILES        Y\n",
              "0  CHEMBL3741945  NC(C1=NNC(=O)c2ccccc12)c3cccc(c3)C(=O)N4CCN(CC... -6.69897\n",
              "1  CHEMBL1682668  COC(=O)NC1(CCN(C[C@@H]2C[C@@]2(C(=O)N(C)Cc3ccc... -6.39794\n",
              "2   CHEMBL213330  Cc1ccccc1CN(CCN(Cc2cncn2C)c3ccc(cc3)c4ccccc4)S... -6.39794\n",
              "3   CHEMBL214610  Cc1ccccc1CN(CCN(Cc2cncn2C)c3ccc(cc3)C#N)S(=O)(... -6.39794\n",
              "4  CHEMBL1277852  CN(CC1=C2C(=CC(=O)N1Cc3ccn(C)n3)NN(C2=O)c4cccc... -6.30103"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-8761bf67-625a-4c15-8352-e09b22c1e2e0\" class=\"colab-df-container\">\n",
              "    <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>Molecule ID</th>\n",
              "      <th>SMILES</th>\n",
              "      <th>Y</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>CHEMBL3741945</td>\n",
              "      <td>NC(C1=NNC(=O)c2ccccc12)c3cccc(c3)C(=O)N4CCN(CC...</td>\n",
              "      <td>-6.69897</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>CHEMBL1682668</td>\n",
              "      <td>COC(=O)NC1(CCN(C[C@@H]2C[C@@]2(C(=O)N(C)Cc3ccc...</td>\n",
              "      <td>-6.39794</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>CHEMBL213330</td>\n",
              "      <td>Cc1ccccc1CN(CCN(Cc2cncn2C)c3ccc(cc3)c4ccccc4)S...</td>\n",
              "      <td>-6.39794</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>CHEMBL214610</td>\n",
              "      <td>Cc1ccccc1CN(CCN(Cc2cncn2C)c3ccc(cc3)C#N)S(=O)(...</td>\n",
              "      <td>-6.39794</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>CHEMBL1277852</td>\n",
              "      <td>CN(CC1=C2C(=CC(=O)N1Cc3ccn(C)n3)NN(C2=O)c4cccc...</td>\n",
              "      <td>-6.30103</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8761bf67-625a-4c15-8352-e09b22c1e2e0')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-8761bf67-625a-4c15-8352-e09b22c1e2e0 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-8761bf67-625a-4c15-8352-e09b22c1e2e0');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-18f6258a-396c-4079-8b21-958972eee072\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-18f6258a-396c-4079-8b21-958972eee072')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-18f6258a-396c-4079-8b21-958972eee072 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_df = df_test.dropna() # random state is a seed value\n",
        "\n",
        "len_test_df=test_df.shape[0]\n",
        "test_df.reset_index(drop=True, inplace=True)\n",
        "\n",
        "adjacency_tensor_test, feature_tensor_test, Y_tensor_test = [], [], []\n",
        "\n",
        "\n",
        "for idx in range(len_test_df):\n",
        "    adjacency, features = smiles_to_graph(test_df.loc[idx][\"SMILES\"])\n",
        "    Y= test_df.loc[idx][\"Y\"]\n",
        "    adjacency_tensor_test.append(adjacency)\n",
        "    feature_tensor_test.append(features)\n",
        "    Y_tensor_test.append(Y)\n",
        "\n",
        "# defining the input for the test set\n",
        "adjacency_tensor_test= np.array(adjacency_tensor_test)\n",
        "feature_tensor_test = np.array(feature_tensor_test)\n",
        "Y_tensor_test= np.array(Y_tensor_test)"
      ],
      "metadata": {
        "id": "mDtrv-gbL3QH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Rescaling the Y tensor\n",
        "Y_tensor_test_rescaled=scaler.fit_transform(Y_tensor_test.reshape(-1, 1))"
      ],
      "metadata": {
        "id": "8n3IPeNGL5Ot"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# checking the performance of the test set\n",
        "y_pred_test = model.predict([adjacency_tensor_test, feature_tensor_test])[2]\n",
        "\n",
        "from sklearn.metrics import r2_score\n",
        "\n",
        "r2 = r2_score(Y_tensor_test_rescaled, y_pred_test)\n",
        "print(\"R-squared value:\", r2)"
      ],
      "metadata": {
        "id": "AnbAoeJYL6n-",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8286ab18-7907-40c3-dda5-1e724171f29e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "38/38 [==============================] - 0s 8ms/step\n",
            "R-squared value: -0.6246354664751446\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.scatter(Y_tensor_test_rescaled,y_pred_test,color='blue', edgecolor='black')\n",
        "plt.xlabel('Actual values rescaled')\n",
        "plt.ylabel('Predicted values')\n",
        "\n",
        "x = np.linspace(plt.xlim()[0], plt.xlim()[1], 100)\n",
        "plt.plot(x, x, label='y=x line', color='black',linestyle='dashed')"
      ],
      "metadata": {
        "id": "zGyoI2lfL8UD",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "outputId": "f3f82205-cd5e-4352-b805-5d1ed323797a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7c816725c880>]"
            ]
          },
          "metadata": {},
          "execution_count": 50
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Here we try to visualize the molecular prediction from the latent dimension. Since the model's performance is not quite up to the task, this prediction is just a proof of principle that if we are able to optimize the earlier model better, then the workflow can be adapted for generation of new molecules with desired log(Papp) values"
      ],
      "metadata": {
        "id": "fUKTpnhIMA3w"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "molecules = new_model.inference(1000)\n",
        "\n",
        "MolsToGridImage(\n",
        "    [m for m in molecules if m is not None][:1000], molsPerRow=5, subImgSize=(260, 160)\n",
        ")"
      ],
      "metadata": {
        "id": "ph4r8T_7L-JF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "c957b627-b32f-4205-d0df-49694558dba2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "32/32 [==============================] - 0s 4ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Display latent space clusters with respect to molecular properties\n",
        "\n",
        "def plot_latent(vae, data, labels):\n",
        "    # display a 2D plot of the property in the latent space\n",
        "    z_mean, _ = vae.encoder.predict(data)\n",
        "    plt.figure(figsize=(12, 10))\n",
        "    plt.scatter(z_mean[:, 0], z_mean[:, 1], c=labels)\n",
        "    plt.colorbar()\n",
        "    plt.xlabel(\"z[0]\")\n",
        "    plt.ylabel(\"z[1]\")\n",
        "    plt.show()\n",
        "\n",
        "\n",
        "plot_latent(new_model, [adjacency_tensor_new[:], feature_tensor_new[:]], Y_tensor[:])"
      ],
      "metadata": {
        "id": "UX4r2F3MMB5f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 867
        },
        "outputId": "e1dcb190-75fc-495a-b0cb-f5ecc129525b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "22/22 [==============================] - 0s 4ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x1000 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_true_internal = Y_tensor_rescaled #  Y_tensor\n",
        "y_pred_internal = new_model.predict([adjacency_tensor_new, feature_tensor_new])[2]\n",
        "\n",
        "y_true_external = Y_tensor_test_rescaled\n",
        "y_pred_external = model.predict([adjacency_tensor_test, feature_tensor_test])[2]\n",
        "\n",
        "# Calculating metrics for Internal Model\n",
        "mae_internal = mean_absolute_error(y_true_internal, y_pred_internal)\n",
        "rmse_internal = mean_squared_error(y_true_internal, y_pred_internal, squared=False)\n",
        "r2_internal = r2_score(y_true_internal, y_pred_internal)\n",
        "medae_internal = median_absolute_error(y_true_internal, y_pred_internal)\n",
        "\n",
        "# Calculating metrics for External Model\n",
        "mae_external = mean_absolute_error(y_true_external, y_pred_external)\n",
        "rmse_external = mean_squared_error(y_true_external, y_pred_external, squared=False)\n",
        "r2_external = r2_score(y_true_external, y_pred_external)\n",
        "medae_external = median_absolute_error(y_true_external, y_pred_external)\n",
        "\n",
        "# Prepare Data for Plotting\n",
        "models = ['Internal Set', 'External Set']\n",
        "mae_values = [mae_internal, mae_external]\n",
        "rmse_values = [rmse_internal, rmse_external]\n",
        "r2_values = [r2_internal, r2_external]\n",
        "medae_values = [medae_internal, medae_external]\n",
        "\n",
        "\n",
        "# Set up the figure and subplots\n",
        "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 6))\n",
        "\n",
        "# Bar width\n",
        "bar_width = 0.2\n",
        "index = np.arange(len(models))\n",
        "\n",
        "# Function to add value labels\n",
        "def add_value_labels(ax, spacing=5):\n",
        "    for rect in ax.patches:\n",
        "        y_value = rect.get_height()\n",
        "        x_value = rect.get_x() + rect.get_width() / 2\n",
        "\n",
        "        label = \"{:.2f}\".format(y_value)\n",
        "        ax.annotate(label, (x_value, y_value), xytext=(0, spacing),\n",
        "                    textcoords=\"offset points\", ha='center', va='bottom')\n",
        "\n",
        "# Plotting MAE, RMSE, and MedAE on the left subplot (ax1)\n",
        "ax1.bar(index, mae_values, width=bar_width, label='MAE', color='blue')\n",
        "ax1.bar(index + bar_width, rmse_values, width=bar_width, label='RMSE', color='orange')\n",
        "ax1.bar(index + 2 * bar_width, medae_values, width=bar_width, label='MedAE', color='red')\n",
        "\n",
        "ax1.set_xlabel('Models')\n",
        "ax1.set_ylabel('Performance Metrics')\n",
        "ax1.set_title('MAE, RMSE, and MedAE')\n",
        "ax1.set_xticks(index + bar_width, models, rotation=45, ha='right', fontsize=8)\n",
        "ax1.legend()\n",
        "\n",
        "# Add value labels to the left subplot\n",
        "add_value_labels(ax1)\n",
        "\n",
        "# Plotting R² on the right subplot (ax2)\n",
        "ax2.bar(index, r2_values, width=bar_width, color='green', label='R²')\n",
        "\n",
        "ax2.set_xlabel('Models')\n",
        "ax2.set_ylabel('R² Value')\n",
        "ax2.set_title('R² Metric')\n",
        "ax2.set_xticks(index, models, rotation=45, ha='right', fontsize=8)\n",
        "ax2.legend()\n",
        "\n",
        "# Add value labels to the right subplot\n",
        "add_value_labels(ax2)\n",
        "\n",
        "# Adjust layout\n",
        "plt.tight_layout()\n",
        "\n",
        "# Display the plot\n",
        "plt.show()\n",
        "\n",
        "\n",
        "# Printing Metrics\n",
        "print(f'MAE: {mae_values}')\n",
        "print(f'RMSE: {rmse_values}')\n",
        "print(f'R²: {r2_values}')\n",
        "print(f'MedAE: {medae_values}')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 711
        },
        "id": "SMvTyMppdSdt",
        "outputId": "de534962-9b21-4540-ba4e-f87978b9a65b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "22/22 [==============================] - 0s 8ms/step\n",
            "38/38 [==============================] - 0s 7ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x600 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MAE: [11.15883319018232, 0.1885973974951963]\n",
            "RMSE: [11.2159671838294, 0.22299486566562635]\n",
            "R²: [-5846.4600645666205, -0.6246354664751446]\n",
            "MedAE: [11.278219765211464, 0.18255801923388293]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Discussion\n",
        "\n",
        "\n",
        "In this notebook, we implemented a VAE-MLP model based on the Chem VAE paper by Gomez-Bombarelli et al., 2018 https://pubs.acs.org/doi/10.1021/acscentsci.7b00572 , with modifications that include features extracted from graphical representations instead of SMILES. We first observed that the performance of the hybrid graphical-based VAE-MLP model on a vast dataset like the Zinc database, comprising 250k molecules is not optimal, evidenced by a consistently poor R² value and significant errors during training.We attempted to evaluate the model's performance by adjusting various hyperparameters, but the results remained qualitatively similar across different configurations.\n",
        "\n",
        "Subsequently, we adapted the model for our dataset in two ways: with and without transfer learning. The initial results using transfer learning involved freezing the encoder and decoder layers and training only the MLP part. However, the range of the predicted values significantly differed from the actual values. This discrepancy might stem from the difference in the nature of the QED data, on which the larger model was trained, and our dataset. To address this, we applied a scaler to the permeability values to align their range with that of the Zinc database. Histograms indicated a similar distribution between the scaled datasets, each exhibiting a skew on one side. Nonetheless, this adjustment did not substantially improve the model's performance.\n",
        "\n",
        "Despite our dataset being considerably smaller than that used in the original Chem-VAE paper (~1,000 vs. ~100,000, 100 fold difference), we evaluated the model's performance without transfer learning, relying solely on our training data. The results show that the model performs better, albeit not as well as simpler models like Random Forest and Graph Convolutional Network, also explored in this notebook. We present metrics demonstrating the model's performance on both our training dataset and an external dataset. Interestingly, the model performed better on the external test set than on the training set, suggesting potential undertraining or uniformly poor performance across both sets. This performance without tranfer learning also suggests that there is some scope of improvement in our approach if the training on the large dataset is done well and with a predictor qualitatively similar to log(Papp) values.\n",
        "\n",
        "Finally, we explored the molecular generation aspect by sampling from the latent dimension. While most generated structures resembled propane, some predictions included basic groups like -NH2 and -OH. We also visualized the latent space in two dimensions. Ideally, molecules should cluster based on their permeability values. However, we observed some variance, though points with lower permeability values tended to cluster towards one end of the graph.\n",
        "\n"
      ],
      "metadata": {
        "id": "cUtJU7z5wQ8T"
      }
    }
  ]
}