Figure 1.

A schematic diagram of the pipeline. CNVrd2 is a modified version of CNVrd. CNVrd2 is identical to CNVrd at the counting, transforming, standardizing and segmenting steps (A,B,G: black text). However, CNVrd2 has additional steps: identification of polymorphic regions (C), merging sub-regions inside genes/regions being measured and testing boundary regions (D), using a simple linear regression model to adjust segmentation scores between populations (E) and a Bayesian normal mixture to cluster segmentation scores of highly CN variable regions into different groups (F). These new steps are in blue text.