Fig. 1

Identification of candidate enhancer-gene pairs under different selection pressures between ancestries. a From left to right: Chromatin accessibility, activity, and contact frequency were assayed in LCLs from four African populations (blue) and four European populations (orange). Enhancer-gene pairs were defined using ABC scores and ATAC, ChIP, and HiC score components quantified. Enhancer-gene pair 1 exemplifies evidence of directional selection (low within- and high between-continental ancestry score variance). b Sequencing reads’ contributions to each ABC component score are shown for a hypothetical enhancer-gene pair where sample 2 has higher ATAC and ChIP scores, but an equivalent HiC score relative to sample 1. Dotted lines extending to the depicted region border connect to reads whose other paired-end aligns elsewhere in the genome but contributes to the HiChIP score (purple gradient). c Principal component analysis results from all enhancer-gene pairs are shown for each score type. Both replicates are shown for each population. CEU is an outlier (Additional file 1: Fig. S5–8) and is thus excluded in these PCAs and in downstream analyses. Abbreviations: ATAC-seq, assay for transpose accessible chromatin with sequencing; E, enhancer; G, gene; H3K27ac, histone 3 lysine 27 acetylation; LCLs, lymphoblastoid cell lines; CEU, Utah residents with North European ancestry; ESN, Esan; FIN, Finnish; GWD, Gambian; IBS, Iberian; LWK, Luhya; TSI, Tuscan; YRI, Yoruban; AFR, African; EUR, European; PC, principal component