omicverse.space.crop_space_visium

omicverse.space.crop_space_visium#

omicverse.space.crop_space_visium(adata, crop_loc, crop_area, library_id, scale, spatial_key='spatial', res='hires', coordinate_order=None, library_key=None)[source]#

Crop Visium spatial data to a specific region of interest.

This function allows cropping of Visium spatial transcriptomics data to focus on a specific region while maintaining proper scaling and coordinate systems.

Parameters:

  • adata – AnnData Annotated data matrix containing Visium spatial data.

  • crop_loc – tuple (x, y) coordinates for the top-left corner of the crop region.

  • crop_area – tuple (width, height) of the cropping area in spatial coordinates.

  • library_id – str Library ID for the spatial data in adata.uns[‘spatial’].

  • scale – float Scale factor for the cropping operation.

  • spatial_key (default: 'spatial') – str, optional (default=’spatial’) Key in adata.obsm containing spatial coordinates.

  • res (default: 'hires') – str, optional (default=’hires’) Image resolution to use (‘hires’ or ‘lowres’).

  • coordinate_order (default: None) – {‘xy’, ‘yx’} or None, default=None Interpretation of crop_loc and crop_area. 'xy' means (x, y) and (width, height). 'yx' preserves the legacy Squidpy-style (y, x) / (height, width) convention. None selects legacy 'yx'.

  • library_key (default: None) – str or None Observation column identifying library membership. Required for a multi-library AnnData.

Returns:

AnnData

Cropped AnnData object containing only spots within the specified region. The spatial coordinates and image are adjusted accordingly.

Notes

  • The function preserves the original coordinate system scaling

  • The cropped image is stored in adata.uns[‘spatial’][library_id][‘images’][res]

  • Coordinates are automatically adjusted to the new cropped region

Examples

>>> import scanpy as sc
>>> import omicverse as ov
>>> # Load Visium data
>>> adata = sc.read_visium(...)
>>> # Crop a 1000x1000 region starting at (500, 500)
>>> adata_cropped = ov.space.crop_space_visium(
...     adata,
...     crop_loc=(500, 500),
...     crop_area=(1000, 1000),
...     library_id='V1_Human_Brain',
...     scale=1.0,
...     coordinate_order='xy'
... )
Contents