Fig. 2

Epigenomic profiles of risk variants and benign variants. a Epigenomic profiles of active chromatin marks, H3K27ac and H3K4me1, repressive chromatin marks, H3K9me3 and H3K27me3, open chromatin regions in the CD14 and K562 cell lines in the neighborhoods of a risk variant, rs3024505 (chr1:206939904), associated with type 1 diabetes, and a benign variant, rs114490664 (chr1:968345). b Distribution of read counts for FAIRE-seq in the K562 cell line across 147 risk variants associated with type 1 diabetes and corresponding benign variants; distribution of read counts of H3K9me3 ChIP-seq in the CD14 cell line across 147 risk variants associated with type 1 diabetes and corresponding benign variants. c Heatmap of standardized read counts of top 100 epigenomic features and bottom 100 epigenomic features across 147 risk variants associated with type 1 diabetes and 147 corresponding benign risk variants. Epigenomic features are ranked by the t-statistics from the most enriched to the most depleted in risk variants compared to benign variants. Read counts are standardized by subtracting the average of read counts of each feature and divided by the standard deviation of read counts of each feature. d Distribution of t-statistics for three types of epigenomic features: TF binding, histone modification, and open chromatin. Within the informative features, 33 informative open chromatin-associated features are enriched while 17 informative open chromatin-associated features are depleted; 96 TF-associated informative features are enriched while 26 TF-associated informative features are depleted; 145 informative histone-associated features are enriched while 187 informative histone-associated features are depleted