Even after enrichment by selection of an appropriate discovery cohort, and testing only for rare GSVs with dominant phenotypic effects, low statistical power remains an important issue, with increased rates of false negative and false positive associations at a given significance threshold and inflated estimates of effect size (“winner’s curse”) [20], [31], [32]. As noted above, increasing cohort size to give improved power and/or reduced type I error rate may not be readily achievable for highly specific phenotypes. A second statistical issue, common to all scans for variant associations, is that each of the many GSVs detected in even a small cohort is the subject of a separate independent hypothesis, so that substantial multiple testing correction of the significance threshold or rigorous control of the false discovery rate (FDR) is required if a large type I error rate is to be avoided. Even after the use of predefined criteria to select a subset of GSVs for analysis, the necessary correction remains substantial: In the previous studies, subjects with common obesity carried approximately 20 GSV calls per individual (1080 cases, 2500
