spcoral.pp.preprogress_adata

spcoral.pp.preprogress_adata#

spcoral.pp.preprogress_adata(adata_omics1, adata_omics2, method='knn', k=10, use_obsm='spatial')#

Preprocess two spatial omics AnnData objects by removing spots/cells with no cross-omics spatial neighbors.

Parameters:

  • adata_omics1 (anndata.AnnData) – First spatial omics AnnData object. Must contain spatial coordinates in obsm[use_obsm].

  • adata_omics2 (anndata.AnnData) – Second spatial omics AnnData object with the same spatial coordinate requirement.

  • method ({'knn'}, optional (default: 'knn')) – Method used to define spatial neighbors across datasets. Currently only ‘knn’ is supported.

  • k (int, optional (default: 10)) – Number of nearest neighbors in the other omics to consider when building the cross-omics adjacency. Each spot in omics1 finds its k nearest spots in omics2, and vice versa (resulting in a bipartite graph).

  • use_obsm (str, optional (default: 'spatial')) – Key in .obsm where spatial coordinates (n_obs × 2) are stored.

Returns:

  • Filtered copy of adata_omics1 containing only spots with at least one neighbor in omics2.

  • Filtered copy of adata_omics2 containing only spots with at least one neighbor in omics1.

Return type:

tuple of (anndata.AnnData, anndata.AnnData)

Contents