Giter Club home page Giter Club logo

ecpi's Introduction

Predicting The Biological Effects of Chemical Protein Interactions

Overview

This code contain scripts and data for experiments on the prediction of biological effects of chemical-protein interactions.

Requirements

This repository only require the installation of tensorflow (standard or gpu) and the LibKGE library.

Usage

You can train a KGE models on the CTD38E dataset by running its corresponding training script from the src directory as follows:

cd src
python experiment_TriVec.py

The script will then run and evaluate the model as follows:

Importing dataset files ...
Initializing the knowledge graph embedding model...
Training ...
2019-12-03 14:39:28,722 - TriModel - DEBUG - Logging model parameters ...
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] _predict_pipeline_on: False
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] batch_size          : 5000
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] em_size             : 100
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] initialiser         : xavier_uniform
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] log_interval        : 10
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] loss                : default
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] lr                  : 0.01
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] nb_ents             : 38
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] nb_epochs           : 100
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] nb_negs             : 2
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] nb_rels             : 38
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] optimiser           : AMSgrad
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] predict_batch_size  : 40000
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] reg_wt              : 0.01
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] seed                : 1234
2019-12-03 14:39:28,722 - TriModel - DEBUG - [Parameter] verbose             : 2
2019-12-03 14:39:28,722 - TriModel - DEBUG - Model training started ...
2019-12-03 14:39:28,722 - TriModel - DEBUG - Training model [ 639965 #Instances - 38 #Entities - 38 #Relations ]
2019-12-03 14:39:30,139 - TriModel - DEBUG - Initialising tensorflow session
2019-12-03 14:39:31,320 - TriModel - DEBUG - Executing tensorflow global variable initialiser
2019-12-03 14:39:32,399 - TriModel - DEBUG - [Training] Epoch # 1    - Speed: 608.626 (k. record/sec) - Loss: 0.6925 - Avg(Loss): 0.6925 - Std(Loss): 0.0004
2019-12-03 14:39:41,246 - TriModel - DEBUG - [Training] Epoch # 10   - Speed: 644.637 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0003
2019-12-03 14:39:51,172 - TriModel - DEBUG - [Training] Epoch # 20   - Speed: 643.619 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0003
2019-12-03 14:40:01,135 - TriModel - DEBUG - [Training] Epoch # 30   - Speed: 644.524 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:40:11,075 - TriModel - DEBUG - [Training] Epoch # 40   - Speed: 643.351 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:40:20,948 - TriModel - DEBUG - [Training] Epoch # 50   - Speed: 642.839 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:40:30,787 - TriModel - DEBUG - [Training] Epoch # 60   - Speed: 644.784 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:40:40,600 - TriModel - DEBUG - [Training] Epoch # 70   - Speed: 648.451 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:40:50,458 - TriModel - DEBUG - [Training] Epoch # 80   - Speed: 646.412 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:41:00,302 - TriModel - DEBUG - [Training] Epoch # 90   - Speed: 666.808 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:41:10,182 - TriModel - DEBUG - [Training] Epoch # 100  - Speed: 655.118 (k. record/sec) - Loss: 0.6922 - Avg(Loss): 0.6922 - Std(Loss): 0.0002
2019-12-03 14:41:10,185 - TriModel - DEBUG - [Reporting] Finished (100 Epochs) - Avg(Speed): 647.857 (k. record/sec) - Avg(Loss): 0.6922 - Std(Loss): 0.0002
[1] N1:AUROC 1.0000 - N1:AUPR 1.0000    N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 165]    REL:increases__O-linked_glycosylation
[2] N1:AUROC 0.9968 - N1:AUPR 0.9968    N10:AUROC 0.9994 - N10:AUPR 0.9936      N50:AUROC 0.9995 - N50:AUPR 0.9750      [Count: 312]    REL:affects__localization
[3] N1:AUROC 1.0000 - N1:AUPR 1.0000    N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 651]    REL:increases__chemical_synthesis
[4] N1:AUROC 1.0000 - N1:AUPR 1.0000    N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 284]    REL:decreases__phosphorylation
[5] N1:AUROC 1.0000 - N1:AUPR 1.0000    N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 18600]  REL:decreases__expression
[6] N1:AUROC 1.0000 - N1:AUPR 1.0000    N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 367]    REL:increases__response_to_substance
[7] N1:AUROC 1.0000 - N1:AUPR 1.0000    N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 542]    REL:affects__reaction
[8] N1:AUROC 0.2993 - N1:AUPR 0.5880    N10:AUROC 0.3102 - N10:AUPR 0.1266      N50:AUROC 0.3069 - N50:AUPR 0.0280      [Count: 137]    REL:increases__uptake
[9] N1:AUROC 1.0000 - N1:AUPR 1.0000    N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 359]    REL:decreases__response_to_substance
[10] N1:AUROC 0.9083 - N1:AUPR 0.9160   N10:AUROC 0.9028 - N10:AUPR 0.5070      N50:AUROC 0.8767 - N50:AUPR 0.1396      [Count: 109]    REL:increases__degradation
[11] N1:AUROC 0.4754 - N1:AUPR 0.6559   N10:AUROC 0.4656 - N10:AUPR 0.1576      N50:AUROC 0.4961 - N50:AUPR 0.0382      [Count: 61]     REL:increases__localization
[12] N1:AUROC 0.9767 - N1:AUPR 0.9773   N10:AUROC 0.9798 - N10:AUPR 0.8323      N50:AUROC 0.9809 - N50:AUPR 0.5109      [Count: 258]    REL:increases__abundance
[13] N1:AUROC 0.9937 - N1:AUPR 0.9937   N10:AUROC 0.9924 - N10:AUPR 0.9296      N50:AUROC 0.9926 - N50:AUPR 0.7287      [Count: 317]    REL:increases__cleavage
[14] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 829]    REL:decreases__methylation
[15] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 1996]   REL:affects__binding
[16] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 50]     REL:affects__splicing
[17] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 199]    REL:increases__metabolic_processing
[18] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 0.9992 - N50:AUPR 0.9615      [Count: 50]     REL:increases__reduction
[19] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 102]    REL:affects__activity
[20] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 570]    REL:affects__response_to_substance
[21] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 72]     REL:increases__mutagenesis
[22] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 432]    REL:affects__methylation
[23] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 1162]   REL:increases__methylation
[24] N1:AUROC 0.8961 - N1:AUPR 0.9059   N10:AUROC 0.9143 - N10:AUPR 0.5385      N50:AUROC 0.9169 - N50:AUPR 0.1940      [Count: 77]     REL:increases__glucuronidation
[25] N1:AUROC 0.0266 - N1:AUPR 0.5067   N10:AUROC 0.0185 - N10:AUPR 0.0925      N50:AUROC 0.0186 - N50:AUPR 0.0200      [Count: 753]    REL:increases__phosphorylation
[26] N1:AUROC 0.0187 - N1:AUPR 0.5047   N10:AUROC 0.0299 - N10:AUPR 0.0934      N50:AUROC 0.0290 - N50:AUPR 0.0202      [Count: 107]    REL:decreases__secretion
[27] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 0.9989 - N50:AUPR 0.9495      [Count: 94]     REL:affects__metabolic_processing
[28] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 78]     REL:increases__oxidation
[29] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 101]    REL:increases__transport
[30] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 9094]   REL:affects__cotreatment
[31] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 4109]   REL:decreases__reaction
[32] N1:AUROC 0.9993 - N1:AUPR 0.9994   N10:AUROC 0.9993 - N10:AUPR 0.9981      N50:AUROC 0.9993 - N50:AUPR 0.9976      [Count: 1172]   REL:decreases__activity
[33] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 1483]   REL:increases__reaction
[34] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 59]     REL:decreases__abundance
[35] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 3778]   REL:affects__expression
[36] N1:AUROC 0.9994 - N1:AUPR 0.9994   N10:AUROC 0.9987 - N10:AUPR 0.9872      N50:AUROC 0.9993 - N50:AUPR 0.9650      [Count: 1544]   REL:increases__activity
[37] N1:AUROC 0.0979 - N1:AUPR 0.5257   N10:AUROC 0.0823 - N10:AUPR 0.0983      N50:AUROC 0.0784 - N50:AUPR 0.0212      [Count: 429]    REL:increases__secretion
[38] N1:AUROC 1.0000 - N1:AUPR 1.0000   N10:AUROC 1.0000 - N10:AUPR 1.0000      N50:AUROC 1.0000 - N50:AUPR 1.0000      [Count: 20585]  REL:increases__expression
-----------------------------------------------------------------------------------------------------------
N1:AUROC 0.8865 - N1:AUPR 0.9360        N10:AUROC 0.8867 - N10:AUPR 0.8514      N50:AUROC 0.8866 - N50:AUPR 0.8039 = [AVERAGE]

ecpi's People

Contributors

samehkamaleldin avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.