Modeling combinatorial regulation from single-cell multi-omics provides regulatory units underpinning cell type...

Area of Science:

• Computational Biology
• Genomics
• Systems Biology

Background:

• Single-cell technologies generate vast omics data, crucial for understanding gene regulatory networks (GRNs).
• Existing GRN inference methods often overlook reusable regulatory modules essential for cell type specificity.

Purpose of the Study:

• To introduce cRegulon, a novel computational approach for inferring regulatory modules from single-cell multi-omics data.
• To model combinatorial regulation by transcription factors (TFs) within diverse GRNs.

Main Methods:

• Inferred regulatory modules by modeling TF combinatorial regulation.
• Utilized diverse GRNs derived from single-cell multi-omics data.
• Benchmarked cRegulon against existing methods using simulated and real datasets.

Main Results:

• cRegulon demonstrated superior performance in identifying TF combinatorial modules as functional regulatory units.
• The method improved cell type annotation accuracy compared to existing approaches.
• Identified universal and reusable regulatory modules underpinning cell type identity.

Conclusions:

• cRegulon provides a robust methodology for dissecting combinatorial gene regulation in single cells.
• The findings offer new insights into the fundamental principles of GRN organization and cell type determination.
• cRegulon advances the analysis of single-cell multi-omics data for biological discovery.