torch_pgn.datasets package

Submodules

torch_pgn.datasets.FPDataset module

class torch_pgn.datasets.FPDataset.FPDataset(args)

Bases: object

Dataset class to load fingerprint data.

process_raw_data(): Load the raw data and format to be compatible with FingerprintDataset

write_fps()

torch_pgn.datasets.ManyVsManyDataset module

class torch_pgn.datasets.ManyVsManyDataset.ManyVsManyDataset(args)

Bases: PGDataset

This dataset is for a 1:1 receptor to ligand mapping. The paths (raw_pdb_path and raw_mol_path) are the same and point to a directory containing subdirectories containing each ligand receptor pair. #TODO: Currently conforms to only pdbbind input formats. Make this much more generic.

process_raw_data()

torch_pgn.datasets.OneVsManyDataset module

class torch_pgn.datasets.OneVsManyDataset.OneVsManyDataset(args)

Bases: PGDataset

This dataset is used to load working_data that is one receptor and many ligands. The pdb is loaded and all mol files are loaded then an pg is constructed for each ligand with the receptor.

process_raw_data()

torch_pgn.datasets.PGDataset module

class torch_pgn.datasets.PGDataset.PGDataset(args)

Bases: ABC

Generic class for loading datasets. It has all of the methods required to write the dataset in order to be compatible with the rest of the processing pipeline.

abstract process_raw_data()

write_graphs()