Association Between Genotypes Of Common And Rare Variants Based On The Cross-validation

Resequencing is a new technique that can identify the rare variants which areassociated with disease. Rare mutations are difficult to tag, since genotypingstudies are designed to detect common variants. However, evidences haveshown that genetic heterogeneity is a probable scenario for common diseases, inwhich multiple rare mutations together explain a large proportion of the geneticbasis for the disease. Next-generation sequencing technology will soon allowsequencing the whole genome of large groups of individuals, and thus will makedirectly testing rare variants possible.We will construct a novel statistical test to test the association betweengenotypes of common and rare variants in a genomic region and a complex trait ofinterest based on cross-validation prediction error. This method aims to test theeffects of both rare and common variants. It is applicable to qualitative traits,allows covariates, can control for population stratification, and is robust todirections of effects of causal variants.Extensive simulation studies show that it is powerful for identifyingdisease-associated genes and more powerful than existing ones either for testingeffects of both rare and common variants or for testing effects of rare variantsonly.