Open AccessMethods for fine-mapping with chromatin and expression data
Author(s) -
Megan Roytman,
Gleb Kichaev,
Alexander Gusev,
Bogdan Paşaniuc
Publication year - 2018
Publication title -
plos genetics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.587
H-Index - 233
eISSN - 1553-7404
pISSN - 1553-7390
DOI - 10.1371/journal.pgen.1007240
Subject(s) - chromatin , epigenome , biology , computational biology , chia pet , genetics , genome , gene expression , heuristic , expression (computer science) , gene , evolutionary biology , chromatin remodeling , computer science , dna methylation , artificial intelligence , programming language
Recent studies have identified thousands of regions in the genome associated with chromatin modifications, which may in turn be affecting gene expression. Existing works have used heuristic methods to investigate the relationships between genome, epigenome, and gene expression, but, to our knowledge, none have explicitly modeled the chain of causality whereby genetic variants impact chromatin, which impacts gene expression. In this work we introduce a new hierarchical fine-mapping framework that integrates information across all three levels of data to better identify the causal variant and chromatin mark that are concordantly influencing gene expression. In simulations we show that our method is more accurate than existing approaches at identifying the causal mark influencing expression. We analyze empirical genetic, chromatin, and gene expression data from 65 African-ancestry and 47 European-ancestry individuals and show that many of the paths prioritized by our method are consistent with the proposed causal model and often lie in likely functional regions.