Computational Enhanced Hi-C data reveals the function of structural geometry in genomic regulation

Abstract

High-throughput chromosome conformation capture (Hi-C) technique profiles the genomic structure in a genome-wide fashion. The reproducibility and consistency of Hi-C data are essential in characterizing dynamics of genomic structures. We developed a diffusion-based method, CTG (Hi-C To Geometry), to deal with the technical bias induced by insufficient sampling in sequencing and obtain reliable gemeotric information of the chromatin. CTG properly quantifies dubiously weak or even undetected interactions and produces a consistent and reproducible framework for the 3D genomic structure. CTG allows for a reliable genome-wide insight on the alteration of genomic structures under different cellular conditions and reveals correlations between genomic-proximal genes at both transcriptional and translational levels. Cell-specific correspondence between gene-gene and corresponding protein-protein physical interactions, as well as that with the transcription correlation reveals the coordinated inter-molecular structural and regulatory information passage in the central dogma.