BART-spatial unravels biologically significant transcriptional regulators from spatial omics data

Area of Science:

• Genomics
• Computational Biology
• Systems Biology

Background:

• Transcriptional regulators (TRs) control cell fate by modulating gene expression.
• Spatial omics technologies map genomic features within their physical context.
• Identifying active TRs in spatial data is challenging due to low expression and indirect correlation with mRNA levels.

Purpose of the Study:

• To develop a computational method for inferring functional TRs from spatial omics data.
• To address the limitations of existing tools in handling spatial heterogeneity and low TR expression.
• To enable the decoding of spatially resolved gene regulatory programs.

Main Methods:

• Developed BART-spatial, a computational method integrating spatial variability and pseudo-temporal information.
• Utilized publicly available TR binding profiles.
• Applied the method to diverse spatial omics datasets (10X Visium, Visium HD, Atera, spatial RNA-ATAC-seq).

Main Results:

• BART-spatial consistently outperforms existing methods in identifying active TRs.
• Identified stage-specific TRs and regulators not detectable by expression analysis alone.
• Demonstrated compatibility and cross-validation with spatial epigenomics data.

Conclusions:

• BART-spatial is a powerful and robust tool for inferring functional TRs in spatial omics data.
• The method enhances understanding of cell fate decisions, tissue organization, and disease mechanisms.
• Enables comprehensive analysis of spatially resolved gene regulatory networks.