spatial-stats-neighborhood-enrichment

To compute neighborhood enrichment for spatial transcriptomics data, you need an annotated adata object with 'celltype' labels and 'x','y' coordinates. The Skill applies squidpy permutation testing on a spatial neighbor graph to generate z-score and count matrices.

Neighborhood enrichment z-scores indicate whether cell types are spatially co-localized or depleted. High positive z-scores show significant co-occurrence in tissue neighborhoods, while negative z-scores indicate spatial avoidance, comparing observed interactions to randomized expectations.

Yes, you can use squidpy permutation testing on single-slice spatial datasets. The workflow requires single-slice data with x,y coordinates and celltype labels to assess spatial relationships and characterize microenvironment patterns.

Cell type neighborhood enrichment analysis requires an adata object containing 'celltype' annotations in the obs DataFrame alongside 'x' and 'y' spatial coordinates. This spatial neighbor graph data enables the calculation of observed versus expected cell interactions.

You can define a specific number of permutations for squidpy neighborhood enrichment testing. The permutation count determines the statistical robustness of the z-score matrix, which evaluates cell type co-localization significance across the spatial neighbor graph.

A limitation of permutation testing for spatial neighborhood enrichment is its restriction to single-slice datasets with predefined celltype annotations. It cannot process multiple slices or unannotated spatial data, requiring complete coordinate and label inputs to function correctly.