Accurate and reproducible functional maps in 127 human cell types via 2D genome segmentation

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

The Roadmap Epigenomics Consortium has published whole-genome functional annotation maps in 127 human cell types by integrating data from studies of multiple epigenetic marks. These maps have been widely used for studying gene regulation in cell type-specific contexts and predicting the functional impact of DNA mutations on disease. Here, we present a new map of functional elements produced by applying amethod called IDEAS on the same data. The method has several unique advantages and outperforms existing methods, including that used by the Roadmap Epigenomics Consortium. Using five categories of independent experimental datasets, we compared the IDEAS and Roadmap Epigenomics maps. While the overall concordance between the two maps is high, the maps differ substantially in the prediction details and in their consistency of annotation of a given genomic position across cell types. The annotation from IDEAS is uniformly more accurate than the Roadmap Epigenomics annotation and the improvement is substantial based on several criteria. We further introduce a pipeline that improves the reproducibility of functional annotation maps. Thus, we provide a high-quality map of candidate functional regions across 127 human cell types and compare the quality of different annotation methods in order to facilitate biomedical research in epigenomics.

Original languageEnglish (US)
Pages (from-to)9823-9836
Number of pages14
JournalNucleic acids research
Volume45
Issue number17
DOIs
StatePublished - Sep 1 2017

Fingerprint

Epigenomics
Genome
Biomedical Research
Mutation
DNA
Genes

All Science Journal Classification (ASJC) codes

  • Genetics

Cite this

@article{983aac48af9243d59272e6063375110c,
title = "Accurate and reproducible functional maps in 127 human cell types via 2D genome segmentation",
abstract = "The Roadmap Epigenomics Consortium has published whole-genome functional annotation maps in 127 human cell types by integrating data from studies of multiple epigenetic marks. These maps have been widely used for studying gene regulation in cell type-specific contexts and predicting the functional impact of DNA mutations on disease. Here, we present a new map of functional elements produced by applying amethod called IDEAS on the same data. The method has several unique advantages and outperforms existing methods, including that used by the Roadmap Epigenomics Consortium. Using five categories of independent experimental datasets, we compared the IDEAS and Roadmap Epigenomics maps. While the overall concordance between the two maps is high, the maps differ substantially in the prediction details and in their consistency of annotation of a given genomic position across cell types. The annotation from IDEAS is uniformly more accurate than the Roadmap Epigenomics annotation and the improvement is substantial based on several criteria. We further introduce a pipeline that improves the reproducibility of functional annotation maps. Thus, we provide a high-quality map of candidate functional regions across 127 human cell types and compare the quality of different annotation methods in order to facilitate biomedical research in epigenomics.",
author = "Yu Zhang and Hardison, {Ross C.}",
year = "2017",
month = "9",
day = "1",
doi = "10.1093/nar/gkx659",
language = "English (US)",
volume = "45",
pages = "9823--9836",
journal = "Nucleic Acids Research",
issn = "0305-1048",
publisher = "Oxford University Press",
number = "17",

}

Accurate and reproducible functional maps in 127 human cell types via 2D genome segmentation. / Zhang, Yu; Hardison, Ross C.

In: Nucleic acids research, Vol. 45, No. 17, 01.09.2017, p. 9823-9836.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Accurate and reproducible functional maps in 127 human cell types via 2D genome segmentation

AU - Zhang, Yu

AU - Hardison, Ross C.

PY - 2017/9/1

Y1 - 2017/9/1

N2 - The Roadmap Epigenomics Consortium has published whole-genome functional annotation maps in 127 human cell types by integrating data from studies of multiple epigenetic marks. These maps have been widely used for studying gene regulation in cell type-specific contexts and predicting the functional impact of DNA mutations on disease. Here, we present a new map of functional elements produced by applying amethod called IDEAS on the same data. The method has several unique advantages and outperforms existing methods, including that used by the Roadmap Epigenomics Consortium. Using five categories of independent experimental datasets, we compared the IDEAS and Roadmap Epigenomics maps. While the overall concordance between the two maps is high, the maps differ substantially in the prediction details and in their consistency of annotation of a given genomic position across cell types. The annotation from IDEAS is uniformly more accurate than the Roadmap Epigenomics annotation and the improvement is substantial based on several criteria. We further introduce a pipeline that improves the reproducibility of functional annotation maps. Thus, we provide a high-quality map of candidate functional regions across 127 human cell types and compare the quality of different annotation methods in order to facilitate biomedical research in epigenomics.

AB - The Roadmap Epigenomics Consortium has published whole-genome functional annotation maps in 127 human cell types by integrating data from studies of multiple epigenetic marks. These maps have been widely used for studying gene regulation in cell type-specific contexts and predicting the functional impact of DNA mutations on disease. Here, we present a new map of functional elements produced by applying amethod called IDEAS on the same data. The method has several unique advantages and outperforms existing methods, including that used by the Roadmap Epigenomics Consortium. Using five categories of independent experimental datasets, we compared the IDEAS and Roadmap Epigenomics maps. While the overall concordance between the two maps is high, the maps differ substantially in the prediction details and in their consistency of annotation of a given genomic position across cell types. The annotation from IDEAS is uniformly more accurate than the Roadmap Epigenomics annotation and the improvement is substantial based on several criteria. We further introduce a pipeline that improves the reproducibility of functional annotation maps. Thus, we provide a high-quality map of candidate functional regions across 127 human cell types and compare the quality of different annotation methods in order to facilitate biomedical research in epigenomics.

UR - http://www.scopus.com/inward/record.url?scp=85031683679&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85031683679&partnerID=8YFLogxK

U2 - 10.1093/nar/gkx659

DO - 10.1093/nar/gkx659

M3 - Article

C2 - 28973456

AN - SCOPUS:85031683679

VL - 45

SP - 9823

EP - 9836

JO - Nucleic Acids Research

JF - Nucleic Acids Research

SN - 0305-1048

IS - 17

ER -