Table 1.

Overlap of integration breakpoints with FANCD2-assocated fragile sites.

Dataset	Integration subgroups	Breakpoints or loci per subgroup (n)	Overlap with FANCD2 sites (n)	%	p-Value
CESC	All breakpoints	1299	229	17.6	0.0001
	Single breakpoints	258	39	15.1	0.0044
	Clustered breakpoints	1041	190	18.3	0.0001
	Condensed clustered breakpoints	326	59	18.1	0.0001
	Combined single and condensed breakpoints	584	98	16.8	0.0001
HNSCCa	All breakpoints	118	21	17.8	0.0039
	Single breakpoints	27	2	7.4	0.5011
	Clustered breakpoints	91	19	20.9	0.0009
	Condensed clustered breakpoints	30	5	16.7	0.2293
	Combined single and condensed breakpoints	57	7	12.3	0.3700

CESC and HNSCC integration breakpoints (±50 Kb flank regions) were grouped by the number of breakpoints per integration locus; single indicates one breakpoint and clustered indicates two or more breakpoints. Integration breakpoints from each subgroup were intersected with FANCD2-enriched regions and the frequency of overlap calculated. For the “All breakpoints,” “Single breakpoints (not clustered),” and "Clustered breakpoints" each breakpoint was tested independently for its overlap with FANCD2-enriched regions, regardless of whether it was part of a cluster or not. For the “Condensed clustered breakpoints” subgroup, the region spanning the most 5′- and 3′-breakpoints of an integration locus was used to test for the overlap with FANCD2-enriched regions. For the “Combined single and condensed breakpoints” subgroup, the “Single breakpoints (not clustered),” and “Condensed clustered breakpoints” subgroups were combined for overlap analysis. The data was permutated 10,000 times to create an expected distribution of overlap. Bold font indicates significant p-values.

^aA single integration breakpoint on chromosome Y was excluded from this analysis.