6. Some investigators have interpreted this as evidence that most - if not all - complex traits are not caused by common genetic polymorphisms but by multiple rare ones 7, for which population-based case-control studies have little or no POWER of detection 5. However, most candidate gene studies carried out to date have been poorly designed in terms of case definition, control selection, genetic marker selection, and in particular sample size, and therefore cannot provide evidence for success or failure of their intended objectives either way. The potential for GWA studies has only recently materialised because of reductions in genotyping costs and more sophisticated specifications of the genotyping arrays in terms of SNP numbers and coverage (see next paper in this series). The latest products provide 300,000 – 1 million SNPs, supplemented by selected sets that are hypothesized to be of increased functional importance. Despite scepticism regarding their power to detect the modest effect sizes of common polymorphisms expected to underlie complex traits, examples of replicated findings have started to emerge 8-14.
