Chromatin loop anchors are associated with genome instability in cancer and recombination hotspots in the germline

Table 1 Overlap of LAPs with domain boundaries, recombination HSs and DSB-prone regions

a) Feature	Feature count	Genomic span (Mb)
LAPs	14,737	73.69
Boundaries	40,615	250.89
HSs	32,984	181.79
DSB-prone regions	84,946	34.69
b) Intersection	Number of LAPs	Fraction of LAPs	Expected fraction (circular permutations)	p value
LAPs ∩ Boundaries	6960	0.47	0.15	< 10⁻⁴
LAPs ∩ HSs	2385	0.16	0.12	< 10⁻⁴
LAPs ∩ DSB-prone regions	10,133	0.69	0.14	< 10⁻⁴
c) Intersection	Number of LAPs	Fraction of LAPs	Expected fraction (feature intersection)	p value
LAPs ∩ HSs ∩ DSB-prone regions	1709	0.12	0.11	< 0.001
LAPs ∩ HSs ∩ boundaries	1172	0.08	0.08	< 0.05
LAPs ∩ HSs ∩ boundaries ∩ DSB-prone regions	836	0.06	0.05	< 0.05

a The number of features and their genomic span. b The number of LAPs that overlap with domain boundaries, HSs or DSB-prone regions, as well as the fraction of unique LAPs in each intersection (as a proportion of the total dataset of 14,737 LAPs). The expected fraction (column 4) and the corresponding p value (column 5) were calculated based on 10,000 circular permutations in R. c Columns 2 and 3 as in b. The expected fraction (column 4) was calculated by multiplying the respective fractions shown in column 3 of b. For example, the fraction of LAPs in the intersection “LAPs ∩ HSs ∩ DSB-prone regions” is equal to 0.16 × 0.69 = 0.11. The hypergeometric test was used to assess the statistical significance of overlap (column 5)

ISSN: 1474-760X