set 'code/run_finetune.sh' as run bash file
Summary of the outcome data reported in the article. Please visit [Data for PBCNet] (https://zenodo.org/records/10339407) to download and unzip results in our article.
The trained PBCNet
Note: The nature of pairwise input required for PBCNet results in one sample appearing in multiple sample pairs. Therefore, to reduce the time spent on data processing during training and prediction, we store most of the data as pickle files. Please visit [Data for PBCNet] (https://zenodo.org/records/10339407) to download and unzip data.
The ligands in the FEP1 set on mol2 and sdf formats; the protein and pocket files on mol2 and pdb formats; and the computing results of intermolecular interactions.
The ligands in the FEP2 set on mol2 and sdf formats; the protein and pocket files on mol2 and pdb formats; and the computing results of intermolecular interactions.
The corresponding pickle files of the FEP1 set.
The corresponding pickle files of the FEP2 set.
The model input files (csv files) for finetune operation.
The corresponding pickle files of the Training set.
The model input files (csv files) for tarining.
The ligands in the selection test set on mol2 and sdf formats and the protein and pocket files on mol2 and pdb formats.
The corresponding pickle files of the selection test set.
The file combined with Graph_save.py are used for processing new ligands and proteins into pickle files.