Test For The Association Between Rare Variant And Complex Disease

Complex diseases/common diseases are determined by common genetic variants and rare genetic variants. Over the past decade, genome-wide association studies (GWAS) under the common disease/common variant (CDCV) assumption have identified hundreds of common variants which are associated with common diseases or traits. However, common variants can only explain a small proportion of heritability for most complex diseases or traits. Recently, with the advance in the next generation sequencing technology, researchers start to focus on the rare variants whose minor allele frequency (MAF) is less than1%. For low MAF. it’s important to build a powerful test for detecting the association between diseases and rare variants.Motivated by the Score-Joint and collapsing methods. we propose to separate the rare variants into two groups based on the detrimental effects and the protective effects. Furthermore. we investigated how the estimated value of disease prevalence affects the power. Extensive simulation studies were performed. The simulated data were generated by logistic regression and the SimRare program for evaluation of the proposed method. For each method, we set different odds ratios to get different datasets. We did large simulations by R program.From the results, the type one error rate of the proposed method is reasonable under the given significance levels (0.05and0.01). We also compare the proposed method to Score-Joint and SKAT. The results show that the proposed method is more powerful than Score-joint and SKAT consistently when effects of rare variants have different directions. In addition, it’s not reasonable to set disease prevalence at0.5for case-control study.