As already described, the successes of linkage-based approaches for the elucidation of genetic determinants of mendelian diseases was unfortunately not mirrored to the same extent in common multifactorial traits showing complex inheritance. Many linkage scans were performed predominantly using affected sibling pairs, but only relatively few reproducible loci were demonstrated. The power of linkage studies based on allele sharing by descent among affected relatives was thought to be low when effect sizes attributable to genetic variants were small, and the variants themselves were present at high frequency (Kruglyak 2008). This appeared to be the situation with common diseases, in contrast to mendelian disorders. Some affected individuals will not necessarily possess the risk allele and indeed have developed the disease as a result of other risk factors (including distinct genetic variants), while the relatively high frequency of the allelic variant (in contrast to the vary rare variants characteristically seen in mendelian disorders) allows for occurrence in a family through multiple founders and loss of the inheritance pattern (Kruglyak 2008).
The recognition that multiple genomic loci were likely to be involved in susceptibility to common multifactorial traits due to variants present at relatively high frequency with an individually small magnitude of effect led to proposals that association studies would be a more powerful approach than family-based linkage studies, and the development of the 'common disease, common variant' hypothesis (Lander 1996; Risch and Merikangas 1996). In 1997 Collins, Guyer, and Charkravarti described the power of a systematic cataloguing of DNA sequence variation for such proposed studies with the aim of achieving a genome-wide level of association (Collins et al. 1997). At the time the most likely source of functional variants was proposed to be in coding DNA and a direct approach was advocated to try and catalogue all such variants, anticipating that the causative functional variant would therefore be included in the association study. It was recognized that some functional variants would lie outside coding DNA, although it is only more recently that the extent to which this is true has become apparent and the myriad ways in which genetic diversity can have functional consequences.
In parallel, Collins and colleagues proposed an indirect approach based on establishing dense maps of SNPs and making use of the effects of linkage disequilibrium to indirectly find association with causative variants (Collins et al. 1997). The functional variant did not need to be directly genotyped as linkage disequilibrium allowed association with genotyped SNPs to be observed through linkage with the ungenotyped functional variant. Common sequence variants in the form of SNPs would provide effective markers to resolve such loci but this would require a much denser set of SNPs than was currently available. The power of linkage disequilibrium mapping in resolving specific disease genes had been previously demonstrated in mendelian diseases, including fine mapping of specific genes after conventional linkage analysis and directly mapping disease genes (Section 2.3.3). Indeed the use of linkage disequilibrium mapping for the analysis of complex traits had been proposed as early as 1986 by Lander and Botstein but required the establishment of an informative set of genomic markers (Lander and Botstein 1986). The proposal now was for an immediate large scale effort to catalogue sequence diversity, in particular SNPs, with an emphasis on coding variants.
It was envisaged that this would require a consortium effort to establish a dense map of at least 100 000 SNPs with data to be publically deposited and freely available. In 1999 the SNP Consortium was launched, followed in 2002 by the International HapMap Project (Section 9.2.4), a remarkable collaborative research effort to catalogue SNP diversity across the genome for a number of human populations, capturing patterns of common human sequence variation, and so informing and enabling future genetic studies of common human disease (Manolio et al. 2008). A number of other studies have contributed to generating the dense maps of common SNP diversity now available for genome-wide analysis of sequence diversity -notably from Perlegen Sciences, which in 2005 reported data on 1.6 million SNPs analysed among a panel of 71 unrelated Americans of European, African, and Asian ancestry (Hinds et al. 2005).
Was this article helpful?