Introduction

Research in genetics in the first decade of the twenty-first century has been dominated by the attempt to characterize common variation in the human genome and its impact on complex phenotypes. The decade opened with the announcement of the completion of the first draft(s) of the human genome (Lander et al., 2001; Venter et al., 2001), which provided one reference sequence. An international effort (The HapMap, 2003), analogous to the one that had facilitated this first achievement, was then devoted to the characterization of common variants in different human populations (originally focusing on trios to represent European, Yoruba, Beijing Chinese, and Japanese populations). By 2007, commercial enterprises had developed technologies that allowed hundreds of thousands of single nucleotide polymorphisms (SNPs) to be genotyped in thousands of individuals at reasonable costs: genome-wide association studies (GWAS), first described in Risch and Merikangas (1996), became possible and popular. These studies aim to identify genetic loci that influence complex phenotypes: that is, traits whose genetic underpinning is not ascribable to one, or even a handful, of genes. When very many loci influence a trait, it is reasonable to assume that the effect of any of these might be quite modest, requiring a large sample size for detection. GWAS, which recruit individuals from a population, without need to study relatives, represent a convincing design in this context, and indeed, they have become the method of choice for many groups.