Fig. 3

TOP2B preferentially interacts with open chromatin. a Correlation of TOP2B ChIP-seq and DNase I hypersensitivity sequencing signal densities across mouse chromosome 19 (Spearman rho = 0.86). Gene density and GC content are also shown. b Nucleosome occupancy profiles for all TOP2B and CTCF peaks (All) centered on the peak summit are shown for proximal (≤1 kb from TSS, red lines) and distal peaks (>1 kb from TSS, blue lines). Profiles for TOP2B peaks not overlapping CTCF peaks (no CTCF), CTCF peaks not overlapping TOP2B peaks (no TOP2B), TOP2B and CTCF peaks not overlapping RAD21 peaks (no RAD21) are also shown. c De novo motif discovery using all TOP2B peaks (50 bp upstream/downstream of the summit) and TOP2B peaks not overlapping CTCF peaks. d Genome browser view of example CTCF bound regions that show allele-specific bias towards the C57BL/6J genome (left panel) and A/J genome (right panel) (y-axis, number of allelic reads). e Correlation of C57BL/6J allele frequencies for CTCF (x-axis) versus TOP2B (y-axis) at TOP2B/CTCF co-occupied regions. f TOP2B (blue) and CTCF (red) C57BL/6J allele frequencies (blue) at TOP2B/CTCF co-occupied regions, categorized based on CTCF allelic binding preference (see “Methods”). C57 > A/J indicates sites with CTCF preference for the C57BL/6J allele, A/J > C57 indicates preference for the A/J allele, and C57 ~ A/J indicates sites with no significant allelic-specific bias. The TOP2B allelic frequencies for both C57BL/6J and A/J enriched CTCF binding sites were significantly different than the TOP2B allelic frequencies in C57 ~ A/J category (p <10^–6; one-sided Wilcoxon rank-sum test)