The curse of the missing heritability

Abstract

Since "the case of the missing heritability" was highlighted 5 years ago (Maher, 2008), scientists have been investigating various possible explanations for this issue (Manolio et al., 2009; Slatkin, 2009; Eichler et al., 2010; Zuk et al., 2012). Recently, Bloom et al. (2013) conducted a linkage analysis in a large yeast Saccharomyces cerevisiae cross with high statistical power to map functional quantitative trait loci (QTL) and found that nearly all the additive genetic contribution can be explained by the detected QTL. It is striking that the "old fashioned" linkage analysis can resolve the missing heritability problem arisen in the high-throughput genome-wide association study (GWAS) era. Compared to human population studies, an intercross creates large linkage disequilibrium (LD) blocks that greatly enhance statistical power but also reduce QTL mapping resolution. Simple simulations (Figure 1) indicate that the real sources or architecture of missing heritability will remain undiscovered due to LD. Breaking down LD would provide better resolution but reduce the power. This commentary is raised to emphasize the trade-off between resolution and statistical power in mapping functional loci.© 2013 Shen.