Genome-wide association studies and beyond

Abstract

Genome-wide association studies (GWAS) provide an important avenue for undertaking an agnostic evaluation of the association between common genetic variants and risk of disease. Recent advances in our understanding of human genetic variation and the technology to measure such variation have made GWAS feasible. Over the past few years a multitude of GWAS have identified and replicated many associated variants. These findings are enriching our knowledge about the genetic basis of disease and leading some to advocate using GWA study results for genetic testing. For many of the GWA study results, however, the underlying mechanisms remain unclear and the findings explain only a limited amount of heritability. These issues may be clarified by more detailed investigations, including analyses of less common variants, sequence-level data, and environmental exposures. Such studies should help clarify the potential value of genetic testing to the public's health.

Figure 1

Chromosomal locations of genome-wide association (GWA) study results through March 2009. Results are given for 398 publications with p-values ≤ 5 × 10⁻⁸. Reproduced from the National Human Genome Research Institute’s GWAS Catalog: http://www.genome.gov/gwastudies. Credit: D. Leja and T. Manolio

Figure 2

Top panel: cumulative number of GWA study SNPs reported with p-values <10⁻⁵ over time. Bottom panel: GWA study findings by study sample size and number of SNPs per study over time. Each circle indicates a single publication, and the area of the circle reflects the number of associated SNPs in that study. From these plots we can see the rapid increase in GWA study SNPs and a slight trend toward larger studies and more noteworthy SNPs per study (19).

Figure 3

Top panel: number of cases and controls in each published GWA study; many GWAS have fewer cases than controls. Bottom panel: total GWA study sample size (cases and controls) versus effect size for studies of binary traits; larger studies are better powered to detect smaller odds ratios (19).

Figure 4

Top panel: GWA study p-values and associated SNP minor allele frequencies (MAF). The red line is a smoothed curve across these values and highlights that there is little impact of MAF on p-values for the strongest SNPs from GWAS. Bottom panel: GWA study p-values and corresponding odds ratios. There is a slight trend toward smaller odds ratios having smaller p-values (19).