Genome-wide association studies: theoretical and practical concerns

Key Points Recent improvements in genotyping technology and in our knowledge of human genetic variation have made it possible to carry out genome-wide genetic association studies to identify susceptibility genes for common disease. However, before such studies are undertaken, it is important to know what proportion of human genetic variation they are likely to survey and what the likely costs will be per true-positive result. The allelic spectrum of complex diseases — that is, the range of the frequencies and effect sizes of susceptibility loci —will influence the success of genome-wide association studies. Large sample sizes will be required to search even the most accessible end of this spectrum: that is, alleles with population allele frequencies exceeding 0.1 and with odds ratios of 1.3 and above. The existence of extensive regions of linkage disequilibrium in the human genome will greatly reduce the cost of genotyping in genome-wide association studies, as this allows the use of tag SNPs that provide information on a number of other SNPs that are not directly genotyped. However, as much as 30% of the common variants in the genome remains unknown and the latest high-throughput technologies convert only about 50% of SNPs into robust assays. In addition, extensive resequencing and genotyping will have to be carried out to ensure that SNPs in regions of low linkage disequilibrium are surveyed comprehensively. Other factors that need to be taken into account in the design and execution of initial genome-wide association studies, to avoid loss of statistical power, include possible selection bias, population substructure and misclassification errors. To fully understand the allelic variation that underlies common diseases, complete genome sequencing for many individuals with and without disease is required. This is still not technically feasible. However, recently it has become possible to carry out partial surveys of the genome by genotyping large numbers of common SNPs in genome-wide association studies. Here, we outline the main factors — including models of the allelic architecture of common diseases, sample size, map density and sample-collection biases — that need to be taken into account in order to optimize the cost efficiency of identifying genuine disease-susceptibility loci.