Abstract

Numerous empirical genomic studies and long established genetic theory show that complex traits – including many common diseases – are likely to be polygenic with numerous non-coding variants conferring risk of disease via the regulation of gene expression and post-translational modification. While ever increasing sample size and imputation are presented as the solution to relatively underpowered genome-wide association studies, improved study design (for example to minimise disease and genetic heterogeneity) and the importance of genetic marker coverage continue to be neglected. Here we use genetic maps and multi-marker method of association based upon the Malécot-Morton model to map a total of 173 type 2 diabetes disease (T2D) susceptibility loci in European and African American samples. In addition, we use the same mapping methods and adipose expression data to show that approximately 2/3 of these disease loci appear to act as expression Quantitative Trait Loci regulating neighbouring genes (cis-eQTLs), thereby posing testable molecular mechanisms and pathways for conferring risk of T2D disease.

The use of genetic maps provide the following advantages over single SNP tests of association: 1) additional power by the inclusion of “out-of-sample” high-coverage linkage disequilibrium (LD) information with gene mapping methods being more efficient using markers located upon genetic rather than physical maps; 2) the model is more sensitive than single SNP tests to detect functional variants that are potentially in LD with a combination of neighbouring genotyped markers, but not necessarily in high LD with any single genotyped marker; 3) greater commensurability between different SNP array platforms and a biologically meaningful integration of functional annotation upon genetic maps; 4) the model is highly interpretable, both in terms of population genetic theory and in the context of association mapping by providing precise genetic location estimates for putative functional variant(s) in a region.