spatial-domain-stagate

What problem does it solve?

STAGATE-based spatial domain detection identifies spatially coherent regions in single-slice spatial transcriptomics data by learning spatially informed embeddings that jointly model gene expression and neighborhood structure.

Core Features & Use Cases

• STAGATE-based domain detection using a graph-attention auto-encoder that jointly models gene expression and spatial neighbor relationships.
• Works on cell-level and spot-level data, requires x,y coordinates, and outputs adata.obs['spatial_domain'] with per-cell/spot domain labels.
• Flexible domain counting via n_domains (or auto-detection); supports Gaussian Mixture (tied covariance) or Leiden-based clustering and stores domain metadata for downstream visualization.

Quick Start

Provide a single-slice spatial transcriptomics dataset and execute the STAGATE workflow to generate spatial_domain labels in the AnnData object.