Validation of VN-based processing for sample CRL-2324, Chr.1. A: Traditional paired processing, nine tumor-normal pairs. B: Traditional unpaired processing using the HapMap reference set. The SD is roughly 1.6 times higher than for paired processing, C: VN algorithm based processing (HapMap data set as a template). Note the SD is even less than for paired processing (by ~10%) which can be explained by elimination of the instrument noise in the virtual reference signal, which is essentially a modified average of all tumor samples. The SD is reduced ~ 1.7 times compared to the processing using the HapMap reference set.
Lisovich et al. BMC Medical Genomics 2011 4:14 doi:10.1186/1755-8794-4-14