Inferring spatial and signaling relationships between cells from single cell transcriptomic data.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) offers high resolution of individual cell gene expression but often loses spatial context.
• Reconstructing spatial information is critical for understanding tissue organization and cellular interactions.

Purpose of the Study:

• To present SpaOTsc, a novel method for recovering spatial properties from scRNA-seq data.
• To enable the estimation of spatial gene expression, cell-cell communications, and intercellular gene-gene information flow.

Main Methods:

• SpaOTsc utilizes structured optimal transport to integrate spatial measurements with scRNA-seq data.
• A spatial metric is established by mapping scRNA-seq cells to spatial measurements.
• Partial information decomposition is employed to compute intercellular gene-gene information flow.

Main Results:

• The method was validated on four datasets, demonstrating accurate spatial gene expression prediction.
• SpaOTsc successfully identified known cell-cell communications and spatial gene regulations.
• The approach effectively reconstructs spatial cellular dynamics.

Conclusions:

• SpaOTsc is a powerful tool for integrating spatial and non-spatial single-cell data.
• The method has broad applications in spatial transcriptomics and understanding tissue architecture.
• SpaOTsc facilitates the reconstruction of spatial cellular dynamics in complex biological systems.