data and code for the paper 'Evolution-strengthened knowledge graph enables predicting the targetability and druggability of genes'
target-disease.txt: The interaction between targets and diseases
train-num-TransE_l2.txt: The knowledge graph embedding of targets in train dataset
test-num-TransE_l2.txt: The knowledge graph embedding of targets in test dataset
You can replace the above data files with your own data
If you directly use the target-disease association feature of targets we provide in file targetf.npy:
python GraphEvo.py
If you want to train the target-disease association feature of targets yourself, you can use the following command:
python gcn.py
gcn.py is able to automatically generate target and disease lists from the provided target-disease associations in target-disease.txt, then you can use the following command to train the model.
python GraphEvo.py