We next performed signal intensity adjustment and generated CNV calls by PennCNV (cnvPartition is a built-in algorithm that cannot be applied on adjusted signal intensities). With wave adjustment, the PennCNV algorithm generated 17, 11, 11, 8 and 15 calls, for the five samples with increasing DNA quantity. Despite having fewer CNV calls after adjustment, a higher concordance rate is achieved between five samples (Supplementary Table 4): six CNV regions were consistently detected in all samples, including three CNVs that have identical boundaries in all samples. In comparison, without adjustment, only two and five CNV regions were consistently detected by cnvPartition and PennCNV, respectively. This data set further demonstrates that the wave adjustment procedure reduces spurious CNV calls and increases the concordance rate on duplicated samples.
