kerasdeep_neural_networks_exercise_python.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyPXoNBrgUYJpnEIHOkNQi4u",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Intertangler/ML4biotech/blob/main/kerasdeep_neural_networks_exercise_python.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## import the data\n",
        "This artificial data represents the gene expression levels (normalized already) of two separate genes. The class labels associated with each data point indicate the presence or absence 1 or 0 of a particular downstream phenotype influenced by the genes. Our goal here is to detect a nonlinear relationship between the two gene expression levels that strongly correlates with the downstream phenotype."
      ],
      "metadata": {
        "id": "nRZguxgF_JLU"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "eYcEXOb-tLIW"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "url = \"https://raw.githubusercontent.com/Intertangler/ML4biotech/main/gene_expression_XOR.csv\"\n",
        "df = pd.read_csv(url)\n",
        "\n",
        "X = df.iloc[:, :-1].values\n",
        "y = df.iloc[:, -1].values\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import roc_curve, roc_auc_score\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "np.random.seed(428)\n",
        "\n",
        "# Shuffle data\n",
        "shuffle_idx = np.random.permutation(len(X))\n",
        "X = X[shuffle_idx]\n",
        "y = y[shuffle_idx]\n",
        "\n",
        "# Split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=428)\n",
        "\n",
        "# Train a logistic regression\n",
        "logreg = LogisticRegression()\n",
        "logreg.fit(X_train, y_train)\n",
        "logreg_prob = logreg.predict_proba(X_test)[:, 1]\n",
        "# ROC data for logistic regression\n",
        "logreg_fpr, logreg_tpr, _ = roc_curve(y_test, logreg_prob)\n",
        "logreg_auc = roc_auc_score(y_test, logreg_prob)\n",
        "\n",
        "# Plot data points\n",
        "plt.scatter(X[y == 0][:, 0], X[y == 0][:, 1], label=\"Class 0\")\n",
        "plt.scatter(X[y == 1][:, 0], X[y == 1][:, 1], label=\"Class 1\")\n",
        "plt.xlabel(\"Gene 1\")\n",
        "plt.ylabel(\"Gene 2\")\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "rHeKXSnu6x19"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## exercise\n",
        "Complete the missing lines."
      ],
      "metadata": {
        "id": "njqTB0NY_rVF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense\n",
        "from sklearn.metrics import roc_curve, roc_auc_score\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras.callbacks import LambdaCallback\n",
        "\n",
        "\n",
        "#🌟🌟🌟🌟 YOUR CODE HERE 🌟🌟🌟🌟#  # Initialize the model\n",
        "#🌟🌟🌟🌟 YOUR CODE HERE 🌟🌟🌟🌟# # Add one or more hidden layers with a certain number of nodes\n",
        "#🌟🌟🌟🌟 YOUR CODE HERE 🌟🌟🌟🌟# # Add the output layer with a sigmoid activation\n",
        "#🌟🌟🌟🌟 YOUR CODE HERE 🌟🌟🌟🌟# # Compile the model with binary cross-entropy loss and adaptive moment estimation\n",
        "\n",
        "\n",
        "# Function to print the loss every 100 epochs\n",
        "print_callback = LambdaCallback(\n",
        "    if epoch % 100 == 0 else None\n",
        ")\n",
        "\n",
        "# Training part\n",
        "model.fit(\n",
        "    X_train,\n",
        "    y_train,\n",
        "    epochs=1000,\n",
        "    verbose=0,  # No output\n",
        "    callbacks=[print_callback]  # Print using callback\n",
        ")\n",
        "\n",
        "# predict on test set\n",
        "y_pred_test = model.predict(X_test)\n",
        "mlp_prob_keras = y_pred_test.flatten()\n",
        "#  ROC data\n",
        "keras_mlp_fpr, keras_mlp_tpr, _ = roc_curve(y_test, mlp_prob_keras)\n",
        "keras_mlp_auc = roc_auc_score(y_test, mlp_prob_keras)\n",
        "# plotting part!\n",
        "plt.figure()\n",
        "plt.plot(logreg_fpr, logreg_tpr, label=f'logistic Regression (AUC = {logreg_auc:.2f})')\n",
        "plt.plot(keras_mlp_fpr, keras_mlp_tpr, label=f'Keras MLP (AUC = {keras_mlp_auc:.2f})')\n",
        "plt.plot([0, 1], [0, 1], linestyle='--', label='random performance(AUC = 0.5)')\n",
        "plt.xlabel('false positives ')\n",
        "plt.ylabel('true positives ')\n",
        "plt.legend()\n",
        "plt.show()\n"
      ],
      "metadata": {
        "id": "a-07BC-U63GM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "vMUHqtcifGJc"
      },
      "execution_count": 3,
      "outputs": []
    }
  ]
}