| Copy number variation (CNV) has long been known to play a role in rare human disease such as Down syndrome, chr22 deletion syndrome and a number of nervous system disorders. The impact of CNV on complex disease etiology, however, is largely undetermined, and the contribution of CNV to non-disease genetic diversity also remains uncertain. Many techniques exist to search for CNV in genetic data. Unfortunately, each of these technologies is limited in its ability to assess CNV on a genome-wide scale, either by resolution, time, cost, or some combination thereof. At the same time, genotyping technologies have become relatively cheap, and SNP genotyping studies are now commonplace in genetic research. Addressing the contribution of CNV to disease is complicated as the difficulties involved in CNV discovery are compounded by the problems underlying CNV association testing. Developing methods to test CNV for association with disease was recently described as a pressing need. Our goal was to develop a set of computational tools capable of harnessing the information from SNP genotyping arrays in order to not only discover CNV, in this case deletions specifically, but also to then test the discovered deletions for disease association. To this end, we developed a likelihood-based framework that forms the foundation of our computer programs Microdel and Microdel. v2, which are described in detail here. These programs have been thoroughly evaluated in simulation, and applied on real data sets to assess the contribution of microdeletions to both disease and non-disease genetic diversity. |
