Figure 2.

Mutagenicity of potential non-B DNA regions. (A) Variant densities of different genetic regions compared to randomly selected regions. See Materials and Methods, subsection entitled ‘Genic regions,’ for the detailed definitions of different regions. Variant density was defined as total number of nucleotide variants in a specific region normalized by total length of that region. (B) Variant densities of individual potential non-B DNA regions in different genetic regions compared to randomly selected regions. (C) Variant density of potential non-B DNA regions (not including SIDD) with different conservation levels. The calculation was done based on potential non-B DNA regions in human genome. Potential non-B DNA regions were classified as one of: ‘not aligned,’ ‘aligned but not conserved’ or ‘conserved’ based on genome alignment of human-chimpanzee (see Materials and Methods for details). The neighbor region for a specific non-B DNA region was defined as the upstream and downstream region, with each portion having half the length of the corresponding potential non-B DNA region. (D) Probability density of nucleotide variants as a function of their MAF. The probability density here was defined as number of nucleotide variants with a specific MAF, ranging from 0.05 to 0.5 using 0.01 as the step and smoothed by a window of 0.1 centered at each step, normalized by total number of nucleotide variants. Probability densities for all nucleotide variants (solid curve) and variants only in potential non-B DNA regions (dotted curve, not including SIDD) were shown. The differences between variant densities were evaluated through χ² tests of 2 × 2 tables (see Materials and Methods for details). Differences between probability densities were evaluated using the Kolmogorov–Smirnov test. The significance level was categorized as: ***P-value < 0.001; **P-value < 0.01; *P-value < 0.05.