Skip to main content
Fig. 1 | Genome Biology

Fig. 1

From: Decoding enhancer complexity with machine learning and high-throughput discovery

Fig. 1

Proportion of cis-regulatory elements in animal genomes. Percentage of the mouse, human, and fruit fly genomes occupied by putative cis-regulatory elements based on histone marks, accessible chromatin (from DNase I hypersensitive sites (DHS)), and protein-coding regions (exons; those overlapping predicted regulatory elements are excluded). Upset plots show the log10 number of kbs for each region. Mouse accessible regions were profiled in 55 cell and tissue types and cis-regulatory elements were defined by the analysis of H3K4me1, H3K4me3, and H3K27ac histone marks across multiple tissues [15]. Human cis-regulatory elements were defined based on 18-state ChromHMM chromatin models across 98 epigenomes [16, 17]; these elements were defined based on their combination of histone modification profiles across the genome. Enhancer and promoter regions were defined as the union of multiple states (EnhWk, EnhA1, EnhG1, EnhBiv, EnhA2 and EnhG2 for enhancers; TssBiv, TssFlnk, TssA, TssFlnkD and TssFlnkU for promoters). Fruit fly accessible regions were profiled in five embryonic stages (S5, S9, S10, S11, and S14) [13], where DHS regions were separated into proximal (± 1 kb from TSSs) and distal regions

Back to article page