Extended Data Fig. 6: A molecular model for epimutation hotspots.
(a) Schematic representation showing a possible mechanism of interaction between heterochromatin and gbM-SPMRs and how they affect local epimutation rates. (b) CG and non-CG DRDD targets in MA_suvh/4/5/6 (generation 16, lineage 8) and MA_cmt3 (generation 16, lineage 8) are hypomethylated. (c) Venn diagram showing all-contexts DMR overlaps of drdd hyper-DMRs with hypo-DMRs of MA_suvh4/5/6 and MA_cmt3. Asterisks represent significance levels (*** indicates highly significant overlaps (P = 9.9E-05). (d) RNA-Seq analysis in suvh4/5/6 and cmt3 shows a significant downregulation of DML3, and a similar, non-significant trend, for DML2, DME and ROS1 genes. (e) Distribution of total sequence space (base pair) taken up by redCS-SPMRs. The bottom 10% (low sequence space) and top 10% (high sequence space) fraction of all gbM genes are highlighted (top right panel). Violin plots showing gbM frequency of the top 10% and bottom 10% subsets from the histogram (bottom right panel).
