Deciphering driver regulators of cell fate decisions from single-cell transcriptomics data with CEFCON.

Area of Science:

• Computational Biology
• Genomics
• Systems Biology

Background:

• Single-cell technologies facilitate dynamic cell fate mapping.
• Identifying gene regulatory networks and driver factors in cell fate decisions remains challenging.

Purpose of the Study:

• To present CEFCON, a network-based framework for inferring gene regulatory networks (GRNs) and identifying driver regulators of cell fate decisions from single-cell RNA-sequencing (scRNA-seq) data.
• To model cell fate dynamics using network control theory to reveal critical biological processes linked to cell states.

Main Methods:

• Utilized a graph neural network with an attention mechanism to construct cell-lineage-specific GRNs from scRNA-seq data.
• Applied network control theory to model cell fate dynamics and identify driver regulators and gene modules.
• Conducted extensive benchmarking against baseline methods for GRN and driver regulator identification.

Main Results:

• CEFCON demonstrated superior performance in GRN construction, driver regulator identification, and gene module identification compared to existing methods.
• Applied to mouse hematopoietic stem cell differentiation, CEFCON identified key driver regulators for three developmental lineages.
• The identified regulators provided network control insights into stem cell differentiation.

Conclusions:

• CEFCON is a powerful tool for dissecting the mechanisms underlying cell fate decisions using scRNA-seq data.
• The framework enhances understanding of gene regulatory dynamics and driver factors in cellular differentiation.