cs482_finalproject_SNP_only_4mer.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "cs482-finalproject-SNP-only-4mer",
      "version": "0.3.2",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/kaushiksrini/cs482-finalproject/blob/master/cs482_finalproject_SNP_only_4mer.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "dnu103q_nVGF",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import csv\n",
        "import sklearn.model_selection\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "eAgYvkhQpZ93",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!pip install -U -q PyDrive\n",
        "from pydrive.auth import GoogleAuth\n",
        "from pydrive.drive import GoogleDrive\n",
        "from google.colab import auth\n",
        "from oauth2client.client import GoogleCredentials\n",
        "\n",
        "auth.authenticate_user()\n",
        "gauth = GoogleAuth()\n",
        "gauth.credentials = GoogleCredentials.get_application_default()\n",
        "drive = GoogleDrive(gauth)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "trOBC3ySprlF",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "\"\"\"file_list = drive.ListFile({'q': \"'root' in parents and trashed=false\"}).GetList()\n",
        "for file1 in file_list:\n",
        "  print('title: %s, id: %s' % (file1['title'], file1['id']))\"\"\"\n",
        "\n",
        "nbp = drive.CreateFile({'id': '1H_pbNqLMrLqisLwqDP4SNF5AejrBw6XX'})\n",
        "nbp.GetContentFile('nbp.csv')\n",
        "bp = drive.CreateFile({'id': '1ENXkJyRRy45TK4SntiGrmxY1jF6qTx-r'})\n",
        "bp.GetContentFile('bp.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "rbnft5bDhBp4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "ef1cd625-bd52-46fa-d64e-9ecc85ca869c"
      },
      "cell_type": "code",
      "source": [
        "def char_to_val(char):\n",
        "  if char == 'a':\n",
        "    return 0\n",
        "  if char == 't':\n",
        "    return 1\n",
        "  if char == 'c':\n",
        "    return 2\n",
        "  if char == 'g':\n",
        "    return 3\n",
        "  raise UnexpectedNucleotideError(char)\n",
        "\n",
        "\"\"\"with open('nbp.csv', 'r') as f:\n",
        "  nbp = list(csv.reader(f, delimiter=','))\n",
        "with open('bp.csv', 'r') as f:\n",
        "  bp = list(csv.reader(f, delimiter=','))\"\"\"\n",
        "\n",
        "#nbp = [nbp[i][0] + nbp[i][1] for i in range(len(nbp))]\n",
        "#bp = [bp[i][0] + nbp[i][1] for i in range(len(bp))]\n",
        "\n",
        "def deconstruct(data):\n",
        "  for i in range(len(data)):\n",
        "    data[i] = [char_to_val(x) for x in data[i]]\n",
        "\n",
        "#deconstruct(nbp)\n",
        "#deconstruct(bp)\n",
        "\n",
        "  \n",
        "df_nbp = pd.read_csv('nbp.csv', header=None)\n",
        "print(df_nbp.shape)\n",
        "df_bp = pd.read_csv('bp.csv', header=None)\n",
        "print(df_bp.shape)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(48643, 102)\n",
            "(2699, 102)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "2mjA6eF8JRyt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 151
        },
        "outputId": "fc2d33c6-0c32-4be2-dfe1-efb5fa3935bd"
      },
      "cell_type": "code",
      "source": [
        "def to_categorical(df):\n",
        "  for col in df:\n",
        "    df[col] = df[col].astype('category')\n",
        "    \n",
        "def dont_use_gene(df):\n",
        "  for i, col in enumerate(df):\n",
        "    if i < 96:\n",
        "      df.drop([col], axis=1, inplace=True)\n",
        "    else:\n",
        "      break\n",
        "\n",
        "\n",
        "df_nbp['Brain'] = 1\n",
        "df_bp['Brain'] = 0\n",
        "df_total = df_nbp.append(df_bp)\n",
        "\n",
        "y = df_total['Brain']\n",
        "df_total.drop(['Brain'], axis=1, inplace=True)\n",
        "x = df_total\n",
        "\n",
        "dont_use_gene(x)\n",
        "to_categorical(x)\n",
        "x = pd.get_dummies(x)\n",
        "\n",
        "print(x.shape)\n",
        "print(y.shape)\n",
        "x_train, x_val, y_train, y_val = sklearn.model_selection.train_test_split(x, y, test_size=0.2, random_state=0)\n",
        "x_train, x_test, y_train, y_test = sklearn.model_selection.train_test_split(x_train, y_train, test_size=0.25, random_state=1)\n",
        "print(x_train.shape)\n",
        "print(x_test.shape)\n",
        "print(y_train.shape)\n",
        "print(y_test.shape)\n",
        "print(x_val.shape)\n",
        "print(y_val.shape)"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(51342, 1522)\n",
            "(51342,)\n",
            "(30804, 1522)\n",
            "(10269, 1522)\n",
            "(30804,)\n",
            "(10269,)\n",
            "(10269, 1522)\n",
            "(10269,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "36q5e_QE9KhC",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# y_test.head(100)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ucbRdmhnFiwf",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from keras import optimizers\n",
        "from keras import models\n",
        "from keras import layers\n",
        "from keras import metrics\n",
        "from keras.layers import Dense, Activation"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "u7_CN-QwUPUg",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "model = models.Sequential()\n",
        "model.add(Dense(100, activation='relu', input_dim=1522))\n",
        "model.add(Dense(40, activation='relu'))\n",
        "model.add(Dense(10, activation='relu'))\n",
        "model.add(Dense(1, activation='sigmoid'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "GNljzZO2Wfj5",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "W0cP6rXEWVf4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 370
        },
        "outputId": "eb80d19f-b85d-4cd5-a124-45159bd17244"
      },
      "cell_type": "code",
      "source": [
        "history = model.fit(x_train, y_train, validation_data=(x_val, y_val), epochs=10, batch_size=32)"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 30804 samples, validate on 10269 samples\n",
            "Epoch 1/10\n",
            "30804/30804 [==============================] - 8s 259us/step - loss: 0.0857 - acc: 0.9761 - val_loss: 0.0335 - val_acc: 0.9912\n",
            "Epoch 2/10\n",
            "30804/30804 [==============================] - 8s 246us/step - loss: 0.0184 - acc: 0.9953 - val_loss: 0.0264 - val_acc: 0.9947\n",
            "Epoch 3/10\n",
            "30804/30804 [==============================] - 8s 246us/step - loss: 0.0089 - acc: 0.9976 - val_loss: 0.0270 - val_acc: 0.9964\n",
            "Epoch 4/10\n",
            "30804/30804 [==============================] - 8s 247us/step - loss: 0.0058 - acc: 0.9986 - val_loss: 0.0258 - val_acc: 0.9963\n",
            "Epoch 5/10\n",
            "30804/30804 [==============================] - 8s 246us/step - loss: 0.0048 - acc: 0.9987 - val_loss: 0.0287 - val_acc: 0.9963\n",
            "Epoch 6/10\n",
            "30804/30804 [==============================] - 8s 244us/step - loss: 0.0042 - acc: 0.9989 - val_loss: 0.0283 - val_acc: 0.9963\n",
            "Epoch 7/10\n",
            "30804/30804 [==============================] - 8s 247us/step - loss: 0.0042 - acc: 0.9989 - val_loss: 0.0251 - val_acc: 0.9963\n",
            "Epoch 8/10\n",
            "30804/30804 [==============================] - 8s 246us/step - loss: 0.0040 - acc: 0.9989 - val_loss: 0.0284 - val_acc: 0.9962\n",
            "Epoch 9/10\n",
            "30804/30804 [==============================] - 8s 246us/step - loss: 0.0039 - acc: 0.9989 - val_loss: 0.0317 - val_acc: 0.9964\n",
            "Epoch 10/10\n",
            "30804/30804 [==============================] - 8s 246us/step - loss: 0.0038 - acc: 0.9989 - val_loss: 0.0302 - val_acc: 0.9966\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "rC_1XZS_Xv7K",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "aa663d9d-3554-41f5-a1cc-c178a603abe8"
      },
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "y_preds = model.predict(x_test)\n",
        "print(accuracy_score(y_test, np.round(y_preds)))"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9951309767260688\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "SNOzzhq58bFW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 570
        },
        "outputId": "a11f6042-1be2-4907-aec0-de183d24f420"
      },
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['acc'])\n",
        "plt.plot(history.history['val_acc'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEVCAYAAAARjMm4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f44b66bc550>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEVCAYAAAAPRfkLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f44b6708860>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "G3uMmC_qoyK9",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}