To our knowledge, this study represents the most comprehensive analysis of arrays and algorithms for CNV calling performed to date. The results provide insight into platform and algorithm performance, and the data should be a resource for the community that may be used as a benchmark for further comparisons and algorithm development. Our study differs from previous studies in that we have explored the full size spectrum of CNVs in healthy controls, rather than relying on large chromosomal aberrations or creating bacterial artificial chromosome (BAC) clone spike-in samples. As a result, we think this study provides better benchmarks for research aimed at CNV discovery and association, while still providing important insight for data interpretation in clinical laboratories.
