Extended Data Fig. 2: CHMs are hotspots of DNA methylation maintenance in GG and AG pre-implantation embryos.

a, Line charts showing DNA methylation level from SNP-trackable data (C57BL: maternal, DBA: paternal) around hypermethylated CpG-rich regions in GG and AG embryos. b, Line charts showing H3K9me3 signals from SNP-trackable data around H3K9me3-marked CpG-rich regions in GG and AG embryos. c, d, Scatter plots showing the Pearson’s correlation of allelic DNA methylation level (c) and H3K9me3 signals (d) between GG / AG embryos and strain-hybrid embryos in SNP-trackable CpG-rich 1 kb bins, covering ≥ 100 SNP-trackable reads of corresponding WGBS data. e, f, i, j, Bar plot showing the percentage of highly methylated CpGs (DNA methylation level ≥ 0.5) at 2-cell stage maintaining high methylation status during GG (e) and AG (f) pre-implantation embryogenesis and in maternal (C57BL) (i) and paternal (DBA) (j) allele. g, h, k, l, Boxplots comparing DNA methylation level (left) and H3K9me3 signals (right) at stable CHMs (n = 8,384 / 6,433), stage-specific CHMs (n = 15,171 / 19,082) and highly methylated non-CHM CGIs (n = 731 / 703) in GG (g) / AG (h) embryos, or at SNP-trackable (that is, covering ≥ 20 SNP-trackable reads of corresponding WGBS data) stable CHMs (n = 2,072 / 1,499), stage-specific CHMs (n = 3,474 / 4,300) and highly methylated non-CHM CGIs (n = 94 / 96) in maternal (C57BL) (k) and paternal (DBA) (l) allele. Significance between stable CHMs and highly methylated non-CHM CGIs was evaluated by two-sided Wilcoxon rank sum test, ***: p-value < 0.001, **: p-value < 0.01, *: p-value < 0.05. p-values in each panel (from left to right): 4.5×10−44, 4.8×10−5, 3.3×10−69, < 2.2×10−308, < 2.2×10−308, < 2.2×10−308 (g); 4.3×10−126, 3.5×10−239, < 2.2×10−308, < 2.2×10−308, < 2.2×10−308, 1.0 × 10−255 (h); 0.012, 7.2×10−19, 5.6×10−12, 2.8×10−35 (k); 2.3×10−14, 6.8×10−22, 4.5×10−3, 1.4×10−9 (l). The meaning of boxplots is identical to that in Extended Data Fig. 1.