Relationships between CNVR size and G+C content. We partitioned by CNVR size the Redon et al. aCGH set into four equally populated bins; this data set was chosen, as it provides CNVRs covering the broadest logarithmic range of size variation. Differences in G+C content between each of the two largest bins when compared with either of the two smallest bins are significant (P < 4 × 10−3). Even when discarding segmentally duplicated bases, the two larger CNVR sets exhibit significantly increased G+C content (Nijmegen set, 44.2%; Redon et al. aCGH set, 41.5%; P < 1 × 10−3). For each data set, the median size of CNVs within a CNVR is plotted against the CNVs’ median G+C content (filled circles). Redon et al. aCGH CNVRs were also partitioned into four equally populated bins according to median CNV sizes (open circles).
