Optimal tests for rare variant effects in sequencing association studies.

摘要: SUMMARY With development of massively parallel sequencing technologies, there is a substantial need for developing powerful rare variant association tests. Common approaches include burden and non-burden Burden tests assume all variants in the target region have effects on phenotype same direction similar magnitude. The recently proposed sequence kernel test (SKAT) (Wu, M. C., others, 2011. Rare-variant testing data with SKAT.The American Journal HumanGenetics89,82–93],anextensionoftheC-alphatest(Neale,B.M.,andothers,2011.Testingforan unusual distribution variants. PLoS Genetics 7, 161–165], provides robust that particularly presence protective deleterious null variants, but less than when large number are causal direction. As underlying biological mechanisms unknown practice vary from one gene to another across genome,itisofsubstantialpracticalinteresttodevelopatestthatisoptimalforbothscenarios.Inthispaper, we propose class SKAT as special cases, derive an optimal within this maximizes power. We show outperforms wide range scenarios. results illustrated using simulation studies triglyceride Dallas Heart Study. In addition, derived sample size/power calculation formula new family kernels facilitate designing studies.