bio-spatial-transcriptomics-spatial-statistics

This Skill computes Moran's I for spatial transcriptomics data by building spatial neighbor graphs on AnnData objects using Squidpy and Scanpy, then saving spatial autocorrelation results directly to adata.uns.

Moran's I and Geary's C are both spatial autocorrelation statistics computed by this Skill to identify spatial patterns in tissue sections, with Moran's I measuring global spatial autocorrelation and Geary's C emphasizing local dissimilarities between neighboring spots.

Yes, this Skill computes cell type co-occurrence and neighborhood enrichment on AnnData objects using Squidpy, quantifying how specific cell clusters co-localize within the spatial neighborhood graph of your tissue sections.

Subsetting to highly variable genes (HVGs) before running spatial statistics is recommended for scalability and speed, as this Skill processes AnnData objects and subsetting HVGs reduces computational overhead when computing spatial autocorrelation.

Yes, this Skill supports computing Local Moran's I to identify spatially variable genes, calculating local indicators of spatial association across genes or cell types in spatial transcriptomics data processed through Squidpy and Scanpy.

Computing spatial statistics on large spatial transcriptomics datasets is limited by computational overhead when building spatial neighbor graphs across all genes, which is why subsetting to highly variable genes is recommended for faster co-occurrence and autocorrelation processing.