Algorithms such as event-wise testing (EWT) for CNV detection from whole genome sequence data do not require control sequence data; rather, they rely on deviations in coverage depth from the sample's mean depth (34). Since many factors, including GC content, influence a sample's coverage profile, these methods must attempt to correct for these biases to provide adequate specificity (35). The general procedure for these algorithms is to divide the genome into nonoverlapping bins of equal size and then calculate the mean depth of coverage (DOC) for each bin. After the read depths are corrected for GC and other biases, a segmentation algorithm is used to divide the genome into regions of constant copy number.
