This is the code of paper: HiRAND: A novel semi-supervised deep learning framework for predicting drug response to therapy in cancer.
Python 3.7.3
Please install other pakeages by pip install -r requirements.txt
Runing HiRAND on the drug response data: python train_drug.py
Runing HiRAND on simulation data: python train_simulation.py
Runing HiRAND for 5 times on data: sh submit.sh
train_drug.py: main script for drug reponse prediction.
train_simulation.py: main script for simulation data prediction.
until.py: contains definitions for aample similarity graph construction and etc.
model.py: contains PyTorch model definitions for HiRAND.
generate_simulation_data.py: generates the simulation data.
submit.sh: complete the multiple prediction task using train_drug.py or train_simulation.py.
The drug data is too large to upload. If you need them, please download drug data. (retrieve password:bu2d).
For questions or comments about the code please contact: [email protected]