/deepSIBA_pytorch

A Pytorch Implementation of deepSIBA model

Primary LanguagePython

Graph Convolutional Layers and model of DeepSIBA: Implementation in Pytorch

DeepSIBA: chemical structure-based inference of biological alterations using deep learning

Christos Fotis1(+), Nikolaos Meimetis1+, Antonios Sardis1, Leonidas G.Alexopoulos1,2(*)

1. BioSys Lab, National Technical University of Athens, Athens, Greece.

2. ProtATonce Ltd, Athens, Greece.

(+)Equal contributions

(*)Correspondence to: leo@mail.ntua.gr

Original Github repository of the study:

DeepSIBA: chemical structure-based inference of biological alterations using deep learning
Link: https://github.com/BioSysLab/deepSIBA
C.Fotis1(+), N.Meimetis1+, A.Sardis1, LG. Alexopoulos1,2(*)

DeepSIBA Approach

figure1_fl_02

Clone

# clone the source code on your directory
$ git clone https://github.com/NickMeim/deepSIBA_pytorch.git

Learning directory overview

This directory contains the gcnn layers and the model architecture to implement DeepSIBA in Pytorch.

The NGF layers folders contain the source code to implement the graph convolution layers and the appropriate featurization. The code was adapted from https://github.com/keiserlab/keras-neural-graph-fingerprint.

The utility folder contains the following functions:

  • Dataset loaders and generators
  • A function to evaluate the performance of deepSIBA
  • Custom layer and loss function to implement the Gaussian regression layer.