Fig. 5: SVs and 3D organization of chromatin.
a, Hi-C contact frequency signatures for three SV types (deletions, duplications, inversions). Signatures were defined by subsetting Hi-C contact frequencies in a 600-kb search space around each known SV detected using WGS, centered around the two breakpoints of the SV. The individual Hi-C submatrices for each SV in each sample were then averaged across each of the SV types (top) and compared to the size-matched background, where random regions of the same size as the SVs were visualized using the same approach as for the SVs (bottom). b, Schematic showing how the Hi-C contact frequency matrix was defined with respect to SV breakpoints, which were then averaged for all SVs across samples for a. c, Percentage of SVs versus the size-matched background where both ends were in the same TAD in the samples (n = 76 biologically independent samples). As we demonstrated the effect that SVs can have on the Hi-C interaction data, we instead examined TADs in samples other than the one with the SV to assess if the TAD structure potentially predisposes the formation of SVs. The center line represents the median; the box limits represent the upper and lower quartiles; the whiskers represent 1.5× the IQR. P values were computed using a two-sided, paired Wilcoxon rank-sum test. d, Median-centered log-ratio of Hi-C contact frequency comparing mCRPC samples and benign prostate. TMPRSS2–ERG fusion-positive samples (with breakpoints within five bins of the median) above the diagonal (canonical TMPRSS2–ERG fusion was defined using WGS, all but one consistent with monoallelic deletion, with one complex rearrangement) and negative below it. Green, higher in mCRPC; blue, higher in benign prostate.
