et al. 2012). Sample size determines the number of significant loci discovered (Sullivan et al. 2012), so that meta-analysis of results is the gold standard. Consortium GWA studies for traits such as height and body-mass index now report sample sizes of >100,000 (Berndt et al. 2013; Lango Allen and et al. 2010; Speliotes et al. 2010). Consortia for psychiatric disorders and behavioral traits have also been formed, with sample sizes increasing rapidly to hundreds of thousands (Rietveld et al. 2012; Ripke et al. 2011; Wray et al. 2012), leading to the discovery of novel loci for psychiatric disorders and educational attainment. Thus, large sample sizes are essential for behavioral phenotypes.
