Supplementary Figure 6: H3K36me3 is depleted at hyper-DMRs and DMVs. | Nature Genetics

Supplementary Figure 6: H3K36me3 is depleted at hyper-DMRs and DMVs.

From: Gain-of-function DNMT3A mutations cause microcephalic dwarfism and hypermethylation of Polycomb-regulated regions

Supplementary Figure 6

a, Representative genome browser views of H3K36me3 ChIP-Rx-seq experiment. Few H3K36me3 sequencing reads are detected at FOXA1 and SOX1 DMVs. The region surrounding the housekeeping gene GAPDH is plotted as a positive control for comparison, demonstrating high levels of H3K36me3 over gene bodies. DNA methylation (magenta, scale 0–100%, all tracks), H3K27me3 (green, scale 0–4 scaled read counts per 107 reads, all tracks), H3K4me3 (yellow, scale 0–8 scaled read counts per 107 reads, all tracks) in control (C1, C2) and patient (P1, P2) dermal fibroblasts. DNA methylation, H3K27me3 and H3K4me3 data for SOX1 and FOXA1 (Fig. 3f) are reproduced again here for comparison with H3K36me3 coverage. b, Mean H3K36me3 ChIP-Rx levels at DMVs in control fibroblasts. Normalized read counts are plotted, with all DMVs scaled to the same length. The start of the DMV (0%) to the end of the DMV (100%) is highlighted in gray. The horizontal line indicates no enrichment compared to ChIP input control. c, H3K36me3 peaks are significantly depleted at hypermethylated DMRs. Bar graph, mean percentage of Infinium Methylation probes overlapping H3K36me3 peaks in control (C1, C2) and patient (P1, P2) fibroblasts. Data points, percentage of probes overlapping H3K36me3 peaks, for each respective cell line. All, all probes on the array (n = 403,064 probes). Hyper-DMRs, probes within hypermethylated DMRs (n = 9,085). Error bars, s.d. P value, two-sided Fisher’s exact tests all versus hyper-DMR probes.

Back to article page