Back

BayesCNet: Bayesian inference for cell type-specific regulatory networks leveraging cell type hierarchy in single-cell data

Zhao, F.; Roy, A.; Jin, W.; Peters, L.; Brusko, T.; Lu, Q.; Esser, K.; sun, r. c.; Chen, L. A.

2025-09-17 bioinformatics
10.1101/2025.09.11.675686 bioRxiv
Show abstract

Understanding gene regulatory networks (GRNs) is essential for deciphering biological processes and disease mechanisms. Single-cell multiome technologies now enable joint profiling of chromatin accessibility and gene expression, offering an powerful means to infer cell type-specific GRNs. However, existing methods analyze each cell type independently or aggregate data into pseudo-bulk profiles, limiting their ability to resolve rare populations and capture cellular heterogeneity. We introduce BayesCNet, a Bayesian hierarchical model that jointly infers enhancer-gene linkages across all cell types while leveraging their hierarchical relationships for information sharing. Through extensive simulations, BayesCNet consistently outperforms state-of-the-art methods, with the largest improvements in rare cell types. When applied to real datasets, BayesCNet identifies enhancer-gene linkages with higher accuracy validated by promoter-capture Hi-C data, and reconstructs cell type-specific GRNs that highlight key regulators, demonstrating its power to resolve gene regulatory programs across diverse cell types.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.