Fig. 4

GWAS enrichments for enhancer (Enh) HMM chromatin state regions from consolidated Roadmap Epigenomics consortium data: (A) Shown are significant tissue-specific enrichment results (q value < 0.01, BH correction) for a representative random subset with 40 phenotypes from FORGE2 analysis across the NHGRI/EBI GWAS catalogue (downloaded 2 September 2019, full set shown in Additional File 1: Figure S19). GWAS phenotypes (rows) are clustered using complete linkage clustering (Euclidean distance), while different tissues and cell types (columns) are grouped according to related tissue or cell type categories (e.g. white blood cell categories are placed together, as are fibroblast categories, etc.). Values are row-normalised, with reference lineplot on the left indicating top enrichment value for each category. Clustering reveals grouping of related phenotypes with similar tissue-specific enrichment profiles. We highlighted several groups from each tissue, including phenotypes such as self-reported allergy for blood cells, atrial fibrillation for heart tissue or high-density lipoprotein (HDL) cholesterol levels for liver. (B) Number of shared vs unique (non-shared) tissue-specific enrichment profiles for different GWAS phenotypes across 15 chromatin states. (C) Venn diagram showing number of shared and unique (non-shared) phenotypes between DNase-seq datasets (DNase I hotspots), chromatin states and histone mark broadPeak datasets. The largest category is unique enriched phenotypes for histone mark broadPeaks (204, 34.8%), followed by shared enriched phenotypes across all three categories (160, 27.3%). A zoomable version of this figure can be found at: https://www.easyzoom.com/imageaccess/9e3b9bab699148f586db26ad36ba0002