Datasets
Models
Downloads: 1
Diff of
/DEMO/MPNN_CNN_Kiba.ipynb
[000000]
..
[0a9449]
Switch to side-by-side view
--- a +++ b/DEMO/MPNN_CNN_Kiba.ipynb @@ -0,0 +1,231 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "import os\n", + "os.chdir('../')" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [], + "source": [ + "import DeepPurpose.DTI as models\n", + "from DeepPurpose.utils import *\n", + "from DeepPurpose.dataset import *" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Beginning Processing...\n", + "Beginning to extract zip file...\n", + "Done!\n", + "in total: 118254 drug-target pairs\n", + "encoding drug...\n", + "unique drugs: 2068\n", + "drug encoding finished...\n", + "encoding protein...\n", + "unique target sequence: 229\n", + "protein encoding finished...\n", + "splitting dataset...\n", + "Done.\n", + "cost about 195 seconds\n" + ] + } + ], + "source": [ + "from time import time\n", + "\n", + "t1 = time()\n", + "X_drug, X_target, y = load_process_KIBA('./data/', binary=False)\n", + "\n", + "drug_encoding = 'MPNN'\n", + "target_encoding = 'CNN'\n", + "train, val, test = data_process(X_drug, X_target, y, \n", + " drug_encoding, target_encoding, \n", + " split_method='random',frac=[0.7,0.1,0.2])\n", + "\n", + "# use the parameters setting provided in the paper: https://arxiv.org/abs/1801.10193\n", + "config = generate_config(drug_encoding = drug_encoding, \n", + " target_encoding = target_encoding, \n", + " cls_hidden_dims = [1024,1024,512], \n", + " train_epoch = 100, \n", + " test_every_X_epoch = 10, \n", + " LR = 0.001, \n", + " batch_size = 128,\n", + " hidden_dim_drug = 128,\n", + " mpnn_hidden_size = 128,\n", + " mpnn_depth = 3, \n", + " cnn_target_filters = [32,64,96],\n", + " cnn_target_kernels = [4,8,12]\n", + " )\n", + "model = models.model_initialize(**config)\n", + "t2 = time()\n", + "print(\"cost about \" + str(int(t2-t1)) + \" seconds\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Let's use CPU/s!\n", + "--- Data Preparation ---\n", + "--- Go for Training ---\n", + "Training at Epoch 1 iteration 0 with loss 139.439. Total time 0.00305 hours\n", + "Training at Epoch 1 iteration 100 with loss 0.85528. Total time 0.20138 hours\n", + "Training at Epoch 1 iteration 200 with loss 0.40343. Total time 0.34138 hours\n", + "Training at Epoch 1 iteration 300 with loss 0.84512. Total time 0.47444 hours\n", + "Training at Epoch 1 iteration 400 with loss 0.66289. Total time 0.60805 hours\n", + "Training at Epoch 1 iteration 500 with loss 0.85238. Total time 0.74027 hours\n", + "Training at Epoch 1 iteration 600 with loss 0.81808. Total time 0.87611 hours\n", + "Validation at Epoch 1 , MSE: 0.72030 , Pearson Correlation: 0.43944 with p-value: 0.0 , Concordance Index: 0.68554\n", + "Training at Epoch 2 iteration 0 with loss 1.07556. Total time 0.99277 hours\n", + "Training at Epoch 2 iteration 100 with loss 0.93850. Total time 1.11694 hours\n", + "Training at Epoch 2 iteration 200 with loss 0.50950. Total time 1.24111 hours\n", + "Training at Epoch 2 iteration 300 with loss 0.65378. Total time 1.36888 hours\n", + "Training at Epoch 2 iteration 400 with loss 0.53988. Total time 1.50416 hours\n", + "Training at Epoch 2 iteration 500 with loss 0.88967. Total time 1.63472 hours\n", + "Training at Epoch 2 iteration 600 with loss 0.75896. Total time 1.76277 hours\n", + "Validation at Epoch 2 , MSE: 0.62829 , Pearson Correlation: 0.51199 with p-value: 0.0 , Concordance Index: 0.70848\n", + "Training at Epoch 3 iteration 0 with loss 0.91195. Total time 1.87888 hours\n", + "Training at Epoch 3 iteration 100 with loss 0.87829. Total time 2.01361 hours\n", + "Training at Epoch 3 iteration 200 with loss 0.66226. Total time 2.14527 hours\n", + "Training at Epoch 3 iteration 300 with loss 0.57633. Total time 2.28111 hours\n", + "Training at Epoch 3 iteration 400 with loss 0.97758. Total time 2.4175 hours\n", + "Training at Epoch 3 iteration 500 with loss 0.86774. Total time 2.55361 hours\n", + "Training at Epoch 3 iteration 600 with loss 1.21729. Total time 2.69055 hours\n", + "Validation at Epoch 3 , MSE: 0.50566 , Pearson Correlation: 0.53635 with p-value: 0.0 , Concordance Index: 0.71654\n", + "Training at Epoch 4 iteration 0 with loss 0.62071. Total time 2.80777 hours\n", + "Training at Epoch 4 iteration 100 with loss 0.52476. Total time 2.93527 hours\n", + "Training at Epoch 4 iteration 200 with loss 0.67531. Total time 3.07222 hours\n", + "Training at Epoch 4 iteration 300 with loss 0.55189. Total time 3.20916 hours\n", + "Training at Epoch 4 iteration 400 with loss 0.68485. Total time 3.34583 hours\n", + "Training at Epoch 4 iteration 500 with loss 0.44150. Total time 3.48277 hours\n", + "Training at Epoch 4 iteration 600 with loss 0.47156. Total time 3.61916 hours\n", + "Validation at Epoch 4 , MSE: 0.56851 , Pearson Correlation: 0.56797 with p-value: 0.0 , Concordance Index: 0.72315\n", + "Training at Epoch 5 iteration 0 with loss 0.88668. Total time 3.73333 hours\n", + "Training at Epoch 5 iteration 100 with loss 0.62687. Total time 3.865 hours\n", + "Training at Epoch 5 iteration 200 with loss 0.50301. Total time 4.00138 hours\n", + "Training at Epoch 5 iteration 300 with loss 0.81095. Total time 4.12805 hours\n", + "Training at Epoch 5 iteration 400 with loss 0.67774. Total time 4.26472 hours\n", + "Training at Epoch 5 iteration 500 with loss 0.86882. Total time 4.40138 hours\n", + "Training at Epoch 5 iteration 600 with loss 0.49849. Total time 4.53777 hours\n", + "Validation at Epoch 5 , MSE: 0.48221 , Pearson Correlation: 0.57899 with p-value: 0.0 , Concordance Index: 0.72680\n", + "Training at Epoch 6 iteration 0 with loss 0.61511. Total time 4.65444 hours\n", + "Training at Epoch 6 iteration 100 with loss 0.81947. Total time 4.79166 hours\n", + "Training at Epoch 6 iteration 200 with loss 0.66717. Total time 4.92805 hours\n", + "Training at Epoch 6 iteration 300 with loss 0.63587. Total time 5.06305 hours\n", + "Training at Epoch 6 iteration 400 with loss 0.59426. Total time 5.19916 hours\n", + "Training at Epoch 6 iteration 500 with loss 0.65938. Total time 5.33472 hours\n", + "Training at Epoch 6 iteration 600 with loss 0.55365. Total time 5.47083 hours\n", + "Validation at Epoch 6 , MSE: 0.45493 , Pearson Correlation: 0.59954 with p-value: 0.0 , Concordance Index: 0.73420\n", + "Training at Epoch 7 iteration 0 with loss 0.45920. Total time 5.58888 hours\n", + "Training at Epoch 7 iteration 100 with loss 0.29518. Total time 5.71361 hours\n", + "Training at Epoch 7 iteration 200 with loss 0.65242. Total time 5.84916 hours\n", + "Training at Epoch 7 iteration 300 with loss 0.70349. Total time 5.98722 hours\n", + "Training at Epoch 7 iteration 400 with loss 0.62729. Total time 6.11111 hours\n", + "Training at Epoch 7 iteration 500 with loss 0.50375. Total time 6.23583 hours\n", + "Training at Epoch 7 iteration 600 with loss 0.36792. Total time 6.37027 hours\n", + "Validation at Epoch 7 , MSE: 0.46791 , Pearson Correlation: 0.61566 with p-value: 0.0 , Concordance Index: 0.74448\n", + "Training at Epoch 8 iteration 0 with loss 0.36232. Total time 6.4875 hours\n", + "Training at Epoch 8 iteration 100 with loss 0.51948. Total time 6.62333 hours\n", + "Training at Epoch 8 iteration 200 with loss 0.75429. Total time 6.7575 hours\n", + "Training at Epoch 8 iteration 300 with loss 0.44889. Total time 6.89333 hours\n", + "Training at Epoch 8 iteration 400 with loss 0.49060. Total time 7.02972 hours\n", + "Training at Epoch 8 iteration 500 with loss 0.59408. Total time 7.15944 hours\n", + "Training at Epoch 8 iteration 600 with loss 0.49257. Total time 7.29611 hours\n", + "Validation at Epoch 8 , MSE: 0.42919 , Pearson Correlation: 0.63366 with p-value: 0.0 , Concordance Index: 0.74538\n", + "Training at Epoch 9 iteration 0 with loss 0.49596. Total time 7.41361 hours\n", + "Training at Epoch 9 iteration 100 with loss 0.80841. Total time 7.54583 hours\n", + "Training at Epoch 9 iteration 200 with loss 0.50147. Total time 7.68305 hours\n", + "Training at Epoch 9 iteration 300 with loss 0.61141. Total time 7.81694 hours\n", + "Training at Epoch 9 iteration 400 with loss 0.58063. Total time 7.95277 hours\n", + "Training at Epoch 9 iteration 500 with loss 0.48687. Total time 8.08083 hours\n", + "Training at Epoch 9 iteration 600 with loss 0.64048. Total time 8.21472 hours\n", + "Validation at Epoch 9 , MSE: 0.41064 , Pearson Correlation: 0.64931 with p-value: 0.0 , Concordance Index: 0.75442\n", + "Training at Epoch 10 iteration 0 with loss 0.57446. Total time 8.33111 hours\n", + "Training at Epoch 10 iteration 100 with loss 0.81588. Total time 8.46805 hours\n", + "Training at Epoch 10 iteration 200 with loss 0.59722. Total time 8.605 hours\n", + "Training at Epoch 10 iteration 300 with loss 0.52301. Total time 8.73444 hours\n", + "Training at Epoch 10 iteration 400 with loss 0.58802. Total time 8.86055 hours\n", + "Training at Epoch 10 iteration 500 with loss 0.38357. Total time 8.99666 hours\n", + "Training at Epoch 10 iteration 600 with loss 0.35195. Total time 9.12972 hours\n", + "Validation at Epoch 10 , MSE: 0.39347 , Pearson Correlation: 0.67301 with p-value: 0.0 , Concordance Index: 0.75873\n", + "Training at Epoch 11 iteration 0 with loss 0.37873. Total time 9.24638 hours\n", + "Training at Epoch 11 iteration 100 with loss 0.43624. Total time 9.37888 hours\n", + "Training at Epoch 11 iteration 200 with loss 0.28653. Total time 9.50277 hours\n", + "Training at Epoch 11 iteration 300 with loss 0.35587. Total time 9.6325 hours\n", + "Training at Epoch 11 iteration 400 with loss 0.50246. Total time 9.76861 hours\n", + "Training at Epoch 11 iteration 500 with loss 0.59975. Total time 9.90555 hours\n", + "Training at Epoch 11 iteration 600 with loss 0.40750. Total time 10.0391 hours\n", + "Validation at Epoch 11 , MSE: 0.38795 , Pearson Correlation: 0.69532 with p-value: 0.0 , Concordance Index: 0.76739\n", + "--- Go for Testing ---\n", + "Up to Epoch 10 Testing MSE: 0.48668658921466434 , Pearson Correlation: 0.6011887756161407 with p-value: 0.0 , Concordance Index: 0.7113144222432253\n", + "Training at Epoch 12 iteration 0 with loss 0.46107. Total time 10.2927 hours\n", + "Training at Epoch 12 iteration 100 with loss 0.40440. Total time 10.4205 hours\n", + "Training at Epoch 12 iteration 200 with loss 0.40888. Total time 10.5561 hours\n", + "Training at Epoch 12 iteration 300 with loss 0.45287. Total time 10.6855 hours\n" + ] + } + ], + "source": [ + "model.train(train, val, test)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "model.save_model('./model_MPNN_CNN_Kiba')" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.7.7" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +}