Table 1. Enrichment of structural mutability in hypomethylated regions determined by the germline methylation index (MI = 0) and whole-genome bisulfite sequencing of human sperm DNA (at 2.5× and 15×).

	Enrichment fold (p-value)	Windows with MI = 0	sperm lowest 5%		MI>0 & sperm<5%		MI = 0 & sperm>5%		MI = 0 & sperm<5%
			2.5×	15×	2.5×	15×	2.5×	15×	2.5×	15×
Human evolution	Human-specific rearrangements	10.2 (3.9e-106)	5.1 (2.0e-77)	7.1 (3.7e-226)	3.6 (2.6e-35)	5.2 (1.3e-78)	6.0 (1.3e-28)	2.6 (7.0e-3)	12.5 (5.1e-76)	12.1 (4.0e-134)
Structural polymorphisms	270HapMap CNVs	2.7 (4.2e-13)	1.8 (5.7e-8)	2.3 (1.6e-26)	1.7 (6.5e-5)	2.4 (2.0e-22)	2.5 (2.8e-7)	2.1 (3.2e-3)	3.4 (2.1e-9)	3.2 (1.2e-13)
	450HapMap CNVs	2.0 (1.6e-12)	1.3 (6.0e-4)	1.5 (1.9e-16)	1.3 (6.3e-4)	1.4 (8.6e-8)	1.8 (5.9e-8)	1.3 (1.9e-2)	2.0 (3.1e-6)	2.3 (1.5e-17)
	WTCCC CNVs	2.6 (2.3e-24)	1.6 (1.1e-9)	2.2 (3.0e-56)	1.5 (3.1e-7)	2.2 (1.5e-34)	2.5 (1.8e-17)	1.8 (1.1e-3)	2.5 (6.3e-9)	3.2 (1.3e-31)
	400MGL CNVs	3.1 (4.2e-10)	2.6 (4.1e-6)	2.1 (1.1e-10)	2.5 (4.4e-6)	1.6 (2.3e-12)	3.9 (1.4e-5)	1.7 (6.8e-3)	2.8 (3.8e-3)	2.9 (4.2e-10)
Disease studies	Schizophrenia case-specific rare CNVs	4.0 (2.5e-5)	2.7 (1.2e-2)	2.0 (2.7e-2)	2.6 (4.7e-2)	1.8 (7.3e-2)	3.8 (5.0e-3)	5.3 (3.5e-3)	5.7 (4.5e-3)	3.5 (2.0e-3)
	Autism de novo CNVs in cases	3.9 (1.1e-3)	4.1 (1.3e-4)	3.8 (5.0e-3)	4.5 (1.8e-5)	2.1 (7.3e-5)	3.5 (1.8e-2)	2.7 (1.4e-3)	8.2 (1.7e-5)	8.4 (5.5e-3)
	Bipolar case-specific singleton deletions	1.7 (8.3e-2)	2.3 (1.2e-2)	2.0 (9.9e-2)	1.8 (1.8e-2)	1.1 (9.9e-1)	0.6 (7.3e-1)	0.5 (4.7e-1)	1.4 (1.1e-1)	2.7 (9.3e-3)
	Developmental delay rarecase-specific CNVs	2.3 (2.4e-83)	2.9 (4.5e-302)	3.3 (9.9e-80)	2.5 (1.1e-298)	1.9 (1.6e-190)	3.4(5.8e-231)	4.8 (1.4e-276)	1.6 (1.4e-3)	2.1 (3.2e-3)

P-values are calculated using Chi-square test. Significance of enrichment for hypomethylation in rows marked “Human evolution” and “Structural polymorphisms” was calculated relative to randomly selected windows throughout the genome. For rows marked “Disease studies”, significance of enrichment for hypomethylation was calculated using the following controls: for the schizophrenia study, using control-specific rare CNVs; for the autism study, using inherited rare CNVs found in cases; for the bipolar study, using control-specific singleton deletions. The developmental delay study, significance of enrichment for hypomethylation in windows containing rare (<1% population frequency) CNVs found in cases was established using the CNVs found in control group as controls.