Fig. 1: 5-mCpG detected by nanopore sequencing.
From: The correlation between CpG methylation and gene expression is driven by sequence variants
a, 5-mCpG rates computed across individuals yielded the expected bimodal distribution. b, 5-mCpG rates averaged in 100 bp bins relative to the midposition of chromatin makers assayed in relevant cell types, that is, histone modifications (H3K4me3, H3K27ac, H3K36me3 and H3K9me3) in primary neutrophils, CTCF protein binding sites in primary neutrophils and open chromatin regions in CD4+ T cells obtained from Encode and Roadmap epigenomics project. Additionally, eRNA and main TSS reference maps for RNA samples isolated from whole blood based on cap analysis of gene expression sequencing assays obtained from the Fantom5 project (SlideBase). DHS, DNase hypersensitive sites; eRNA, enhancer RNA. c, We applied sequence-based phasing to assign 5-mCpG status to paternal (p) or maternal (m) haplotypes in each individual (I). For each CpG unit and each parental haplotype, we computed the 5-mCpG rate and defined unmethylated haplotypes where we found three or more neighboring CpG units each with 5-mCpG rate <0.15, but located no more than 500 bp apart, in at least one haplotype (restricted to 2,648 individuals sequenced to an average coverage of >20×). A rectangle is drawn around neighboring CpG units where such unmethylated haplotypes were detected. The cluster labeled α defines locations containing overlapping unmethylated haplotypes labeled a, b and c. The most frequently occurring unmethylated haplotype (fmax) is then nominated as the representative MDS of cluster α (MDSα) containing CpG units x3,x4,x5,x6.
