Trends in Genetics
OpinionNext generation disparities in human genomics: concerns and remedies
Section snippets
The genomics of inequality
With surprisingly little careful evaluation of the potential contributions to health-care disparities 1, 2, the study of human genetic disease has developed an almost exclusive focus on individuals of European ancestry [3]. This bias is illustrated most starkly by genome-wide association studies (GWAS), which have been performed at a ratio of ∼10:1 European ancestry versus all other groups combined (http://www.genome.gov/gwastudies). Additionally, studies of other populations have generally
The generalization of genomics
Modern genomics discovery efforts have therefore been uneven in their sampling of human genomic variation. Whether any clinically actionable genetic discoveries would contribute to health-care disparities depends on both the generality of genetic effects and the generality of the variants themselves. Unfortunately, because many genetic associations have not yet been traced to the causal variant [14], there are only a few cases from which to judge the generality of genetic effects, but some
Discovery genetics using whole-genome sequencing
The most significant difference in strategy between a GWAS and a whole-genome sequencing study is that association shifts from an indirect to a direct framework. That is, GWAS rely on LD between the genotyped and the causal variant to detect the association, whereas in whole-genome sequencing, the causal variants (even if not yet identified) are, by definition, genotyped directly. Thus, the differing patterns of LD are no longer a constraint on discovery. Nevertheless, without corrections,
Identifying unusual variants
One of the discovery strategies for rare variants in complex disease has been to target types of variant that are exceptional because of their infrequency in controls, most notably large heterozygous deletions. As just one example of this approach, Need et al. showed that deletions >2Mb were absent in a cohort of >2500 controls, but present in 0.8% of patients with schizophrenia, suggesting that large deletions are pathogenic [23]. Similarly, Istara et al. found that, in a collection of 2500
Concluding remarks
The shift towards whole-genome sequencing does not remove concerns about uneven discoveries among racial or ethnic groups, but rather emphasizes the importance of those concerns. On balance, we view the GWAS era as having flirted with disaster: had GWAS discoveries led to a series of clinically actionable findings, it is likely that these would have been more applicable to Europeans than to other groups because of the uneven representation of Europeans in these studies. We are among those who
Acknowledgments
We thank Hunt Willard, Charmaine Royal, Misha Angrist and Bob Cook-Deegan for commenting on the manuscript. We also thank the anonymous reviewers for their helpful suggestions, which substantially improved the article.
References (44)
Population analysis of large copy number variants and hotspots of human genetic disease
Am. J. Hum. Genet.
(2009)Ethical and practical challenges surrounding genetic and genomic research in developing countries
Acta Trop.
(2009)- et al.
Will tomorrow's medicines work for everyone?
Nat. Genet.
(2004) - et al.
The A's, G's, C's, and T's of health disparities
BMC Med. Genomics
(2009) - et al.
Merging and emerging cohorts: necessary but not sufficient
Nature
(2007) Genetic homogeneity of Icelanders
Nat. Genet.
(2000)Genomics takes hold in Asia. Interview by David Cyranoski
Nature
(2008)Genomics forges ahead in East Asia
Nat. Biotechnol.
(2000)Principal components analysis corrects for stratification in genome-wide association studies
Nat. Genet.
(2006)A worldwide survey of haplotype variation and linkage disequilibrium in the human genome
Nat. Genet.
(2006)