De novo deciphering three-dimensional chromatin interaction and topological domains by wavelet transformation of epigenetic profiles

Yong Chen, Yunfei Wang, Zhenyu Xuan, Min Chen, Michael Q. Zhang

Research output: Contribution to journalArticlepeer-review

21 Scopus citations

Abstract

Defining chromatin interaction frequencies and topological domains is a great challenge for the annotations of genome structures. Although the chromosome conformation capture (3C) and its derivative methods have been developed for exploring the global interactome, they are limited by high experimental complexity and costs. Here we describe a novel computational method, called CITD, for de novo prediction of the chromatin interaction map by integrating histone modification data. We used the public epigenomic data from human fibroblast IMR90 cell and embryonic stem cell (H1) to develop and test CITD, which can not only successfully reconstruct the chromatin interaction frequencies discovered by the Hi-C technology, but also provide additional novel details of chromosomal organizations. We predicted the chromatin interaction frequencies, topological domains and their states (e.g. active or repressive) for 98 additional cell types from Roadmap Epigenomics and ENCODE projects. A total of 131 protein-coding genes located near 78 preserved boundaries among 100 cell types are found to be significantly enriched in functional categories of the nucleosome organization and chromatin assembly. CITD and its predicted results can be used for complementing the topological domains derived from limited Hi-C data and facilitating the understanding of spatial principles underlying the chromosomal organization.

Original languageEnglish (US)
Article numbere106
JournalNucleic acids research
Volume44
Issue number11
DOIs
StatePublished - Jun 20 2016

ASJC Scopus subject areas

  • Genetics

Fingerprint

Dive into the research topics of 'De novo deciphering three-dimensional chromatin interaction and topological domains by wavelet transformation of epigenetic profiles'. Together they form a unique fingerprint.

Cite this