spatial-domain-detection

To detect spatial domains in spot-level transcriptomics data, you can use SpaGCN to integrate gene expression, spatial coordinates, and optional H&E histology, producing per-spot domain assignments stored in adata.obs['spatial_domain'].

Yes, you can use SpaGCN for multi-slice workflows to compare domain architectures across multiple tissue slices. It computes per-spot spatial domain assignments, enabling cross-slice comparisons of tissue architecture and pathology.

SpaGCN spatial domain detection requires spot-level gene expression data and spatial coordinates as inputs. It also optionally accepts H&E histology features to improve domain calls and ensure accurate adjacency graphs.

Integrating H&E histology improves spatial domain detection in Visium datasets by combining histology features with gene expression and coordinates. This integration helps SpaGCN delineate tissue domains more accurately.

Identifying tissue domains with SpaGCN involves parameter tuning alongside ensuring proper adjacency graphs. The process optimizes the integration of gene expression and coordinates to compute spatial_domain labels for each spot.