Contain the gold standard drug-disease set and ten drug-related networks.
Contain the preprocessing code to generate PPMI matrix.
Contain the PPMI matrices of ten drug-related networks.
- Create two directories "test_models" and "test_results" in the project.
- To get drug features learned by MDA, run
- python getFeatures.py example_params.txt
- To predict drug-disease associations by cVAE, run
- pretraining with features: python cvae.py --dir dataset -a 6 -b 0.1 -m 300 --save --layer 1000 100
- refine training with rating: python cvae.py --dir dataset --rating -a 15 -b 3 -m 500 --load 1 --layer 1000 100
deepDR is tested to work under Python 3.6
The required dependencies for deepDR are Keras, PyTorch, TensorFlow, numpy, scipy, and scikit-learn.