/DRAGNN

Primary LanguagePython

Drug repositioning based on weighted local information augmented graph neural network

This is our tensorflow implementation of DRAGNN for drug repositioning associated with:

Drug repositioning based on weighted local information augmented graph neural network
Yajie Meng, Yi Wang, Junlin Xu, Changcheng Lu, Xianfang Tang, Bengong Zhang, Geng Tian and Jialiang Yang

Environment Requirement

  • tensorflow-gpu == 1.13.1
  • keras == 2.2.4
  • scikit-learn == 0.22.2

Datasets

Usage

  • Please set the mode parameter to "cv," "case," and "analysis" in main.py to reproduce the 10-fold cross-validation results, case study results, and network analysis prediction results reported in our paper, respectively. When mode is specified as "case," please provide the desired "specific_name" to identify the specific case for the analysis.
  • For "case" and "analysis", we strongly recommend running on Fdataset using CPU to ensure accurate reproducibility.

If you find our codes helpful, please kindly cite the following paper. Thanks!

@article{DRAGNN,
  title={Drug repositioning based on weighted local information augmented graph neural network},
  author={Yajie Meng, Yi Wang, Junlin Xu, Changcheng Lu, Xianfang Tang, Bengong Zhang, Geng Tian and Jialiang Yang},
  year={2023},
}