The Impact of Phenotypic and Genetic Heterogeneity on Results of Genome Wide Association Studies of Complex Diseases

摘要: Phenotypic misclassification (between cases) has been shown to reduce the power detect association in genetic studies. However, it is conceivable that complex traits are heterogeneous with respect individual susceptibility and disease pathophysiology, effect of heterogeneity a larger magnitude than phenotyping errors. Although an intuitively clear concept, on studies common diseases received little attention. Here we investigate impact phenotypic statistical genome wide (GWAS). We first performed study simulated genotypic data. Next, analyzed Wellcome Trust Case-Control Consortium (WTCCC) data for diabetes mellitus (DM) type 1 (T1D) 2 (T2D), using varying proportions each order examine strength significance previously found WTCCC In both real data, (presence “non-cases”) reduced greatly decreased estimates risk attributed variation. This finding was also supported by analysis loci validated subsequent large-scale meta-analyses. For example, 50% increases required sample size approximately three times. These results suggest accurate phenotype delineation may be more important detecting true associations increase size.