We have attempted to replicate these recently-reported GSV associations with obesity, using algorithmic analysis of genotyping data from obesity case-control and population cohorts. We replicate association with obesity of GSVs at a single locus on chromosome 16p11.2; this locus is distinct from the association on 16p11.2 which we previously reported (using the same cohorts), being separated from it by >600 kb of intervening sequences; the 2 loci are independently associated with obesity. However, we were unable to replicate a high proportion of the remaining regions, and conclude that there is a need for the development and application of robust statistical methods appropriate for testing for association of rare variants from amongst a large collection of GSVs, independent of the platform used for GSV detection. We also highlight the caution required when attempting to support putative associations by phenotyping affected subjects: phenotypic data from a small number of individuals from a highly-selected cohort may not be reliable as an indication of the impact of the variant in unselected subjects.
