Research On Rare Variant Association Method Based On Multiple Test Fusing

With the emergence of next-generation sequencing technologies, association studies will pay more attention to complex diseases and rare variants. And the common variable correlation test methods have got a better development, but in recent years, studies have shown that rare variants have a big impact on single gene disorders and common diseases. Compared with the association of common variants, rare variants have a lower minor allele frequency. The association test method of common variants in most cases is not suitable for rare variants association study. Therefore, there have been some specifically method for the association test of rare variants. This article mainly studies the correlation between disease and rare variants, including the following two points:For the problem of the robustness of these methods, this paper provides a variety of tests method based on the integration. Several kinds the existing rare variants or common variants association test methods are chosen. A selection criterion is to choose those tests in one or more cases that perform better. Calculating corresponding P values, these methods score for association according to the P values, and then a hypothesis test is for score. This strategy can avoid power losses due to a single test for some sensitive data sets. Primarily through the five simulation experiments dataset to compare with other methods, the experimental results show that the integration proposed a variety of test method to improve the stability of rare variants association test.By merging the grouping and weighted idea can form a grouping weighted sum method. First group the rare variation according to the minor allele frequency, and then weight the rare variants of grouping. The strategy can avoid power losses due to the existing common variants. Experiment on separate rare variants data set, the result show that the method can achieve the best test power on the data set.