Leveraging epigenomes and three-dimensional genome organization for interpreting regulatory variation

Area of Science:

• Genomics
• Systems Biology
• Computational Biology

Background:

• Understanding how regulatory variants impact complex traits is difficult due to unknown gene targets and cell-type contexts.
• Cell-type-specific long-range regulatory interactions are key to studying these impacts, but high-resolution maps are scarce.
• Identifying gene subnetworks targeted by variants remains a challenge.

Purpose of the Study:

• To develop a computational method for predicting cell-type-specific long-range regulatory interactions.
• To create a framework for identifying gene networks targeted by variants from genome-wide association studies (GWAS).
• To interpret regulatory single nucleotide polymorphisms (SNPs) associated with complex phenotypes.

Main Methods:

• Developed L-HiC-Reg, a Random Forests regression model, to predict high-resolution contact counts in various cell types.
• Integrated L-HiC-Reg predictions with a network-based framework to identify variant-targeted gene networks.
• Applied the approach to 55 cell types from the Roadmap Epigenomics Mapping Consortium and analyzed GWAS data.

Main Results:

• Successfully predicted regulatory interactions across 55 cell types, enabling the interpretation of regulatory SNPs.
• Characterized fifteen complex phenotypes, including schizophrenia, coronary artery disease (CAD), and Crohn's disease.
• Identified differentially wired gene subnetworks, revealing both known and novel gene targets of regulatory SNPs.

Conclusions:

• The developed L-HiC-Reg method and network analysis pipeline provide a powerful framework for studying the context-specific effects of regulatory variation.
• This approach enhances the understanding of the genetic architecture of complex diseases by linking regulatory variants to specific gene networks.
• The study offers a valuable compendium of interactions and an analysis pipeline for future research in regulatory genomics and complex trait genetics.