rnaglib.tasks.VirtualScreening¶

class rnaglib.tasks.VirtualScreening(root, ligand_framework='dgl', recompute=False)[source]¶

RNA binding pocket-small molecule binding affinity prediction

Task type: binary classification Task level: substructure-level

Parameters:

root (str) – path to a folder where the task information will be stored for fast loading.
ligand_framework (str) – the package to use to do geometric deep learning on the ligand graph (either “dgl” or “pyg”, default “dgl”)
recompute (bool) – whether to recompute the task info from scratch or use what is stored in root.

Methods

`__init__`(root[, ligand_framework, recompute])
`build_dataset`()
`evaluate`(model)
`get_split_datasets`([dataset_kwargs])	Sets the train, val and test datasets Call this each time you modify `self.dataset`.
`get_split_loaders`([dataset_kwargs, ...])

Attributes