HiCHub: A Network-Based Approach to Identify Domains of Differential Interactions from 3D Genome Data

Abstract

Chromatin architecture is important for gene regulation. Existing algorithms for the identification of interactions changes focus on loops between focal loci. Here we develop a network-based algorithm HiCHub to detect chromatin interaction changes at larger scales. It identifies clusters of genomic elements in physical proximity in one state that exhibit concurrent decreases in interaction among them in the opposite state. The hubs exhibit concordant changes in chromatin state and expression changes, supporting their biological significance. HiCHub works well with data of limited sequencing coverage and facilitates the integration of the one-dimensional epigenetic landscape onto the chromatin architecture. HiCHub provides an approach for finding extended architectural changes and contributes to the connection with transcriptional output. HiCHub is freely available at https://github.com/WeiqunPengLab/HiCHub.