zjshao22 / MFFM

MFFM

files

├── data
│   ├── df_drug.csv //ddie dataset
│   ├── event.db //ddie dataset unzip event.zip
│   ├── processed
│   │   ├── ccatp_feature.pt // chemistry feature concatenate pretrained feature generated by c&p.ipynb
│   │   ├── ddi_drug1.pt // graph feature generated by create_data.py
│   │   └── ddi_drug2.pt // graph feature generated by create_data.py
│   ├── tr572.pt // transformer prtrained feature by char-level
│   └── tr_word_572.pt // transformer prtrained feature by word-level
├── ddi.csv //generated by c&p.ipynb
├── smiles.csv // generated by c&p.ipynb
├── graph.py
├── c&p.ipynb
├── create_data.py
├── readme.md
└── train.py

step

1. run c&p.ipynb get ddi.csv smiles.csv ccatp_feature.pt
2. run create_data.py get ddi_drug1.pt ddi_drug2.pt
3. run train.py

references

https://github.com/YifanDengWHU/DDIMDL
https://github.com/Sinwang404/DeepDDs/tree/master