README.md

GANLDA

The implementation of “GANLDA: Graph attention network for lncRNA-disease associations prediction”, Wei Lan, Ximin Wu, Qingfeng Chen, Wei Peng, Jianxin Wang, Yiping Phoebe Chen. Neurocomputing, 2021. The GAT layer is based on DGL.

Requirement

• Python 3.6

• Numpy

• dgl

• Sklearn

• scipy

• matplotlib

• random

• math

• h5py

• pickle

• torch

• argparse

• itertools

Data

The diseases and lncRNAs association matrix: lncRNA_disease_Associations.h5

The diseases features: disease_Features.h5

The lncRNAs features: lncRNA_Features.h5

The lncRNAs name: lncRNA-name.xlsx

The disease doid: doid.xlsx

Run

The ganlda init program entry: ganlda_init.py

The 10-fold program entry: tenfold.py

The denovo program entry: denovo.py

Obtain the score matrix

If you want to obtain score matrix by GANLDA framework, please run ganlda_init.py directly.