deeporigin.drug_discovery.docking

This module encapsulates methods to run docking and show docking results on Deep Origin

Attributes

LOCAL_BASE module-attribute

LOCAL_BASE = home() / '.deeporigin'

Number module-attribute

Number = float | int

Classes

Docking

Bases: WorkflowStep

class to handle Docking-related tasks within the Complex class.

Objects instantiated here are meant to be used within the Complex class.

Attributes

tool_version class-attribute instance-attribute

tool_version = '0.4.6'

Functions

get_jobs_df

get_jobs_df(
    *,
    pocket_center: Optional[
        tuple[Number, Number, Number] | list[Number]
    ] = None,
    box_size: Optional[
        tuple[Number, Number, Number] | list[Number]
    ] = None
)

search for all jobs that match this protein and ligands in the Job DB, and return a dataframe of the results

Parameters:

Name	Type	Description	Default
pocket_center	Optional[tuple[Number, Number, Number] | list[Number]]	Optional tuple of (x, y, z) coordinates to filter by pocket center	None
box_size	Optional[tuple[Number, Number, Number] | list[Number]]	Optional tuple of (x, y, z) dimensions to filter by box size	None

get_poses

get_poses() -> LigandSet | None

get all docked poses as a LigandSet

get_results

get_results(
    *, file_type: Literal["csv", "sdf"] = "csv"
) -> DataFrame | None | list[str]

return a list of paths to CSV and SDF files that contain the results from docking

Parameters:

Name	Type	Description	Default
file_type	str	"csv" or "sdf". Defaults to "csv".	'csv'

Returns:

Type	Description
DataFrame | None | list[str]	pd.DataFrame | None | list[str]: DataFrame of results, None if no results found, or list of file paths.

run

run(
    *,
    pocket: Optional[Pocket] = None,
    box_size: Optional[
        tuple[Number, Number, Number]
    ] = None,
    pocket_center: Optional[
        tuple[Number, Number, Number]
    ] = None,
    batch_size: Optional[int] = 32,
    n_workers: Optional[int] = None,
    output_dir_path: Optional[str] = None,
    use_parallel: bool = True,
    approve_amount: Optional[int] = None,
    quote: bool = False,
    re_run: bool = False
) -> JobList | None

Run bulk docking on Deep Origin. Ligands will be split into batches based on the batch_size argument, and will run in parallel on Deep Origin clusters.

Parameters:

Name	Type	Description	Default
pocket	Pocket	pocket object. This can be generated using the pocket finder function.	None
box_size	tuple[float, float, float]	box size	None
pocket_center	tuple[float, float, float]	pocket center	None
batch_size	int	batch size. Defaults to 30.	32
n_workers	int	number of workers. Defaults to None.	None
output_dir_path	str	path to output directory. Defaults to None.	None
use_parallel	bool	whether to run jobs in parallel. Defaults to True.	True
approve_amount	int	amount to approve for the jobs. Defaults to None.	None
quote	bool	whether to request a quote for the jobs. Defaults to False.	False
re_run	bool	whether to re-run jobs. Defaults to False.	False

Returns:

Name	Type	Description
JobList	JobList | None	A JobList containing all the created docking jobs.

show_poses

show_poses()

show docked ligands with protein in 3D

show_results

show_results()

show results of bulk Docking run in a table, rendering 2D structures of molecules