SpatialQuery: scalable discovery and molecular characterization of multicellular motifs from spatial omics data

Area of Science:

• Computational biology
• Spatial omics
• Single-cell analysis

Background:

• Spatially resolved single-cell technologies offer in situ cell profiling.
• Computational methods for discovering multicellular spatial patterns and their molecular programs are limited.

Purpose of the Study:

• Introduce SpatialQuery, a framework for identifying cellular motifs (recurrent multicellular co-localization patterns).
• Enable molecular analyses focused on identified motifs, including differential gene expression and covariation analysis.
• Characterize multicellular interactions and functional tissue units beyond pairwise analyses.

Main Methods:

• Developed SpatialQuery, a computational framework for spatial omics data analysis.
• Utilized differential expression and covariation analyses to uncover spatially modulated genes and coordinated expression changes within motifs.
• Applied the framework to spatial transcriptomics and proteomics datasets.

Main Results:

• Uncovered cross-germ-layer signaling in gut tube patterning.
• Identified disease-specific fibrotic and immunosuppressive niches in kidney and colon tissues.
• Revealed regional determinants of motif-associated transcriptional programs in a mouse brain atlas.
• Demonstrated SpatialQuery's light computational footprint for integration into web-based cell atlas portals.

Conclusions:

• SpatialQuery effectively identifies multicellular spatial patterns and their associated molecular programs.
• The framework advances the analysis of spatial omics data, enabling deeper understanding of tissue organization and function.
• SpatialQuery facilitates interactive visualization and exploration of cell atlas data.