Constructing cell-type-specific gene regulatory networks from cell-cell communication and a global gene regulatory network.

Accurately inferring cell-type-specific gene regulatory networks (GRNs) is crucial for understanding cellular heterogeneity, lineage determination, and disease progression mechanisms. Although single-cell RNA sequencing (scRNA-seq) enables high-resolution expression profiling, its inherent sparsity and high noise levels pose challenges to reliable GRN reconstruction. To address these issues, we propose cell type-specific gene regulatory network (CTN), a regulatory inference framework based on bilateral self-representation matrix decomposition, which integrates bulk RNA-seq data with scRNA-seq data to decompose global regulatory relationships and intercellular communication, thereby reconstructing GRNs at the cell type level. CTN demonstrates high inference accuracy and strong performance in functional module identification. In the analysis of real biological processes, CTN reveals the key regulators and functional modules in the differentiation process for different lineages of the hematopoietic system, identifies differentially perturbed regulatory networks for tumor-related cell types, and depicts potential functional mechanisms in the tumor microenvironment.