R2-DDI: Relation-aware Feature Refinement for Drug-drug Interaction Prediction
This repository is the official implementation of R2-DDI: Relation-aware Feature Refinement for Drug-drug Interaction Prediction. The code is originally forked from Fairseq and DVMP.
Requirements and Installation
- PyTorch version == 1.8.0
- PyTorch Geometric version == 1.6.3
- RDKit version == 2020.09.5
You can build the Dockerfile or use the docker image teslazhu/pretrainmol36:latest.
To install the code from source, you should first install rdkit,
conda install -y -c conda-forge rdkit=2020.09.5
Then, install other dependencies, like
git clone https://github.com/linjc16/R2-DDI.git
pip install fairseq
pip uninstall -y fairseq
pip install ninja
python setup.py build_ext --inplace
Getting Started
Dataset
The drugbank dataset can be seen in dataset folder. The twosides dataset can be download here.
Data Preprocessing
We evaluate our models on DrugBank and TwoSides benchmark sets. ddi_zoo/scripts/data_process and ddi_zoo/scripts/twosides/data_process are folders for preprocessing of DrugBank and TwoSides, respectively. To generate the binary data for fairseq, take the transductive setting for DrugBank as an example, run
python ddi_zoo/scripts/data_process/split_trans.py
python ddi_zoo/scripts/data_process/run_process_trans.py
bash ddi_zoo/scripts/data_process/run_binarize_trans.sh
Note that you need to change the file paths accordingly.
Training and Test
All traning and test scripts can be seen in ddi_zoo/scripts. For instance,
bash ddi_zoo/scripts/train_trans/run_gcn_feat_int_cons.sh new_build1 0.01 1e-4 256
bash ddi_zoo/scripts/train_trans/inf_gcn_feat_int_cons.sh new_build1 0.01 1e-4 256
Contact
Please feel free to submit a Github issue if you have any questions or find any bugs. We do not guarantee any support, but will do our best if we can help.