The Research Of Rare Variants Based On The Genome-wide Association Study

The Genome-Wide Association Study(GWAS),which looks for mutations in gene-specific sequences in the human genome and identifies the single Nucleotide Polymorphisms which associated with the disease or trait.Despite these discoveries,a lot of the genetic dedication to the complex traits remains unexplained.For example,a GWAS analysis of type 2 diabetes in 150000 individuals identified 70 loci at genome-wide significance but that explain only 11% of type 2 diabetes heritiability.Several explanations have been proposed for the so-called problem of “missing heritability.” Because GWASs focus on the identification of common variants which have limited effects of complex diseases and rare variants can explain additional disease risk.It is a top priority in Genome-wide association studies to find the rare variants associated with disease from mass gene sequencing data quickly and efficiently.We improve the hypothesis testing method of Sum of Power SSU.We derive the SSU statistic of a single variant when the null hypothesis is established.Based on this statistic and Fisher's P-value we proposed a new method of calculating the weight.Combining the P-value weight method and the improved SPU method,we proposed our own test method.At the same time,we do a large number of simulation studies in this paper to explore the first type of error rate and the power of our method under the situation of the sample size and the situation of the rare variants are associated with a trait with the different directions and exist disturb SNPs and when the rare variants exist linkage disequilibrium.The simulation was generated by logisitc regression for evaluation of the ever has been proposed way.We also have done many grass simulations by R software.Finally,we applied our method to the real Alzheimer's disease data to further verify the effectiveness of this method.From the simulation results,our method can well control the type one error rate even the samples are small.With interferential SNPs,our method is better than the traditional detection and detect with no distractions.In addition,under the situation of the rare variants are associated with a trait with the different direction our method can accurately identify the risk locis which is an useful method.Applying this method to practical scientific research can reveal the genetic pathogenesis of the disease and provide the theoretical basis for the treatment and prevention of complex diseases.