Figure 4. Boundaries containing disease-associated STRs (daSTRs) are characterized by ultra-high density of CpG islands.
(A) Contour density plot depicting the number of CTCF sites and CpG islands in 120 kb bins representing boundaries with normal-length, matched repeats or those depleted of repeat tracts. Points are colored according to density. daSTRs are marked in blue. (B-C) 2×2 contingency table and Fisher’s Exact Test comparing CpG island density at boundaries with normal-length matched repeats versus (B) no repeat tracts and (C) daSTRs. (D-F) Bar plots comparing localization of daSTRs and matched repeats at or not at boundaries for (D) boundaries with >=3 CpG islands, (E) boundaries depleted of CpG islands and (F) CpG islands not at boundaries. (G-I) Bootstrapped distributions of percent daSTRs or matched repeats overlapping boundaries for (G) boundaries with >=3 CpG islands, (H) boundaries depleted of CpG islands and (I) CpG islands not at boundaries. See also Figures S5-S6.