An improved tree-based statistical method for genome-wide association study
In genetic studies, quantitative traits are found possibly associated with genetic data. Due to advanced sequencing technology, many methods have been proposed in genome wide association study (GWAS) to search the single nucleotide polymorphism (SNP) associated with the traits. Currently several methods that account for the evolutionary relatedness among individuals were developed. When comparing with conventional methods without evolutionary relatedness among individuals, tree based methods are found to have better performance when the population structure increases. In this work, we extend a couple of methods in previous studies by varying the magnitude of relatedness. The magnitude of relatedness of the evolutionary history is controlled by an Ornstein-Uhlenbeck (OU) process through its parameters. Our method combines a pertinent process and phylogenetic comparative method where the incorporated evolutionary history is built by SNP data. We perform simulation as well as analyze drosophila longevity data set.