SpatialLeiden

SpatialLeiden integrates spatial information at various steps of Leiden clustering to perform spatially-aware clustering via multiplex modularity optimisation for spatial omics data.

Key Features:

• Spatially-aware Leiden clustering: SpatialLeiden modifies the standard Leiden algorithm by integrating spatial information at multiple steps rather than treating spatial data as non-spatial.
• Multiplex modularity optimisation: The method employs multiplex modularity optimisation as its underlying computational framework for clustering.
• Computational performance: SpatialLeiden achieves high computational performance comparable to state-of-the-art spatial clustering algorithms.

Scientific Applications:

• Spatial omics clustering: SpatialLeiden identifies natural structure in spatial omics data by finding cells with similar molecular phenotype while accounting for spatial context.
• Cell type annotation: Clustering results enable annotation of cell types in spatially resolved single-cell omics datasets.

Methodology:

SpatialLeiden integrates spatial information at various steps of the Leiden clustering algorithm, rendering it into a spatially aware clustering method through multiplex modularity optimisation.