Pure Python spatial transcriptomics analysis.
- Moran's I spatial autocorrelation
- SVG detection (FDR corrected)
- Spatial domain identification (k-means)
- Neighborhood enrichment (permutation)
- Spatial co-expression network
pip install squidpy scanpy matplotlib pandas numpy scipy
python spatial_engine.py- 2688 spots, 2000 HVGs
- 1332 SVGs (padj<0.05), top: Mbp (Moran I=0.79)
- 7 spatial domains, max neighborhood z=71.3
- 704 co-expression edges