Skip to content

A Convolutional Neural Network based model to predict the peptide binding sites in proteins.

Notifications You must be signed in to change notification settings

WafaaWardah/visual

Repository files navigation

Visual

A Convolutional Neural Network based model to predict the peptide binding sites in proteins.
Please cite the associated publication when using these codes.
Citation: W. Wardah, A. Dehzangi and G. Taherzadeh et al., Predicting protein-peptide binding sites with a deep convolutional neural network, Journal of Theoretical Biology, https://doi.org/10.1016/j.jtbi.2020.110278
Environment: The original experiment was run on GeForce GTX 1060 Ti graphics card.

Data Files

The data files contain sample preprocessed protein data (details found in the JTB paper).

  • train_7_set.csv and train_labels.txt
  • val_7_set.csv and val_labels.txt
  • test_7_set.csv and test_labels.txt

Code Files

  • experiment.py (This is the main file. Run this to start running the experiment.)
  • models.py (This contains the model structure and details.)
  • dataset.py (This file loads datasets.)

Guide

Put all these files in one folder and run the experiment.py file. The experiment will continue running iterations until you manually stop/kill it.

About

A Convolutional Neural Network based model to predict the peptide binding sites in proteins.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages