Revealing spatial multimodal heterogeneity in tissues with SpaTrio

Area of Science:

• Computational biology
• Spatial transcriptomics
• Multi-omics data integration

Background:

• Accurate spatial multi-omics data capture is limited by current single-cell sequencing technologies.
• Understanding tissue architecture and cellular interactions requires integrated spatial multi-modal information.

Purpose of the Study:

• To develop a computational method, SpaTrio, for integrating single-cell multi-omics and spatial transcriptomics data.
• To enable the analysis of gene regulation, cellular interactions, and spatial heterogeneity within tissues.

Main Methods:

• Probabilistic alignment of single-cell multi-omics and spatial transcriptomics datasets.
• Benchmarking using simulation datasets to assess accuracy and robustness.
• Evaluation on biological datasets to identify topological patterns and spatial distributions.

Main Results:

• SpaTrio accurately integrates diverse datasets to reconstruct spatial multi-omics information.
• The method successfully detects topological patterns of cells and molecular modalities.
• SpaTrio uncovers spatially multimodal heterogeneity and spatiotemporal gene regulation.

Conclusions:

• SpaTrio provides a robust computational framework for spatial multi-omics data integration.
• The method enables comprehensive analysis of cellular communication and gene expression in spatial contexts.
• SpaTrio offers valuable multimodal insights for advancing spatial biology research.