Inferring Cell-Cell Communications from Spatially Resolved Transcriptomics Data Using a Bayesian Tweedie Model

Area of Science:

• Computational Biology
• Systems Biology
• Genomics

Background:

• Cellular communication is vital for biological functions, but current single-cell sequencing lacks spatial context.
• Existing methods for analyzing intercellular communication often lack statistical rigor and spatial resolution.

Purpose of the Study:

• To develop a statistically robust method for inferring cell-cell communication from spatially resolved transcriptomics data.
• To address limitations of existing methods by incorporating spatial information and providing directional communication insights.

Main Methods:

• Proposed a generalized linear regression model, BAyesian Tweedie modeling of COMmunications (BATCOM).
• Utilized spatially resolved transcriptomics data, particularly spot-based data.
• Estimated communication scores considering cell type distances and ligand-receptor interactions.

Main Results:

• BATCOM provides accurate and reliable inference of cell-cell communication.
• The model naturally determines the direction of communication between cell types.
• Simulation studies and real-data application demonstrated BATCOM's superior performance compared to existing algorithms.

Conclusions:

• BATCOM offers an innovative solution for inferring cell-cell communication from spatial transcriptomics.
• The method fills critical gaps in understanding biological mechanisms by providing directional and interaction-specific communication insights.
• BATCOM delivers robust and straightforward results for complex biological systems.