HGC-GAN

Predicting lncRNA-disease associations based on heterogeneous graph convolutional generative adversarial network.

Author: Zhonghao Lu.

Overview

This repository contains codes necessary to run the HGC-GAN algorithm.

Running Environment

• Windows environment, Python 3.7
• PyTorch >= 1.10.1

Datasets

All datasets are available at Data.

The details of the data are shown in the table.

	lncRNA	Disease	miRNA	LDA	MDA	LMA
Dataset1	861	432	673	4518	4189	2105
Dataset2	1363	501	1190	5338	6763	2291
Dataset3	240	412	495	2697	13562	1002
Dataset4	1723	236	675	1151	4634	10102

Model

• model.py: the core model proposed in the paper.
• main.py: the main program in the project. Run the entire project by running main.py.
• get_adj.py: run the program to obtain the correlation matrix.
• get_k_mer_feat.py: run the program to obtain the sequence features of lncRNA.
• hetero_graph.py: construct and save heterogeneous graphs.
• train.py: training Model.

Question

• If you have any problems or find mistakes in this code, please contact with me: Zhonghao Lu: [email protected]