Efficiently predicting Drug Target Interaction
To install the TDC package, open the terminal and type:
pip install PyTDC
In the data folder, we provide all three processed datasets used in our model: BindingDB_Kd, KIBA, and DUDE. Besides drug smiles and protein sequence, we provide protein pdb as well as contactmap.
- Download my_data.zip: https://drive.google.com/file/d/1fz8EZBOKFplIBugsX2eueR65ZHBBgEre/view?usp=sharing
- Create a new folder my_data under data
- Unzip my_data.zip in my_data
You should see the following data structure:
--data
----my_data
------my_data
--------bindingdb
--------dude
--------kiba
To train on kd, kiba or dude, simply run
bash train.sh [dataset], dataset is one of kd, kiba or dude
For example, train on kd, you should run:
bash train.sh kd
To view the logs while training, run bash view.sh [dataset]
bash view.sh kd