Statistics and Its Interface
Volume 10 (2017)
Testing for gene-gene interaction in case-control GWAS
Pages: 267 – 277
Detecting gene-gene interaction is an important but challenging task in genome-wide association studies (GWASs). To this end, many statistical methods have been proposed in the literature. However, powerful yet robust approaches are yet to be developed. In this paper we study the gene-gene interaction tests for case-control GWASs. A number of powerful tests can be constructed for given situations. We also discuss some tests for the main effects and the overall tests for association between genotype and phenotype. A simulation study is conducted to compare some of the proposed tests with existing methods. A real data application is also conducted to illustrate the use of the proposed tests.
asymptotically independent, combining p-values, odds ratio, single nucleotide polymorphism