Prediction of lncRNA-disease associations via graph convolutional network based on multi-graph structures and attention mechanism

The Code, Datasets directories contain model codes and tested data sets, respectively. The depended python packages are listed in requirements.txt. The package versions should be followed by users in their environments to achieve the supposed performance.

python 3.7.13

numpy==1.21.6

matplotlib==3.5.3

torch==1.6.0+cu102

torch-geometric==1.6.1

networkx==2.6.3

pandas==1.3.5

scipy ==1.7.3

scikit-learn==0.23.2

The program is in Python 3.7.13 using [Pytorch] backends. Use the below bash command to run MGSGCN.

    python Main.py -d dataset

(If using this code , please cite our paper.)

MGSGCN: Prediction of lncRNA-disease associations by graph convolutional network based on multi-graph structures and attention mechanism