7 What Have We Learned from GWAS?
175
perform the genotyping in a cost-efficient way and, second, approaches which allow
the systematic conditional analysis on specific genetic variants using summary data,
without the need to share individualized genotypes, prohibited due to patient privacy
protections (Yang et al. 2012). Systematic application of these studies to hundreds of
genetic loci identified by GWAS promises to discover additional independent alleles
contributing to disease (explaining additional missing heritability) and a graded
structure of high- to low-risk haplotypes contributing to risk. These haplotypes will
then allow specific candidates for causality to emerge, which can then be tested in
functional or experimental systems.
7.4.4 Genetic Studies of Complex Traits across Ethnicities
Despite the higher prevalence and resulting disproportionate public health burden,
genetic studies of complex disease in non-European populations have not taken
off nearly at the rate of their European counterparts (Bustamante et al. 2011).
Fortunately, and recently, this trend appears to be shifting, for important scientific reasons. Well-powered GWAS for a range of disease traits across a range
of ethnic groups (e.g., Asian, Indian, African-American, Latino) have recently
been published (Saxena et al. 2012; SIGMA Type 2 Diabetes Consortium 2013;
DIAGRAM Consortium 2014) with many more underway. Many of these studies
are also actively working to pool data together and perform meta-analysis and
fine-mapping efforts (DIAGRAM Consortium 2014), forming collaborations with
existing groups who have focused on studies primarily in European populations.
These efforts are very important, as studies across populations worldwide can
identify new genetic risk factors that are more common or operate at stronger
effects across populations (Rosenberg et al. 2010; Pulit et al. 2010), and can
help improve the quantification of genetic risk prediction across ethnicities. A
potentially useful application of data across ethnicities is to exploit different patterns
of LD across groups to aid in fine-mapping efforts that seek to identify causal
alleles for disease. While this approach does have challenges, the combination
of the well-powered primary stages, custom-array genotyping technologies, and
a catalog of genetic variation down to low frequencies identified by the 1000
Genomes Project (1000 Genomes Project Consortium 2012) will test this hypothesis
very directly in the immediate future. Furthermore, at genomic sites where an
association has been established beyond a reasonable doubt, it is not unreasonable
to imagine that additional variants statistically independent of the established
association might exist and may segregate differently across ethnic groups, particularly if they are low-frequency or private to specific ethnicities. Thus, these
data may help not only to triangulate on the gene causal for disease but also
to help uncover the genetic mechanisms underlying disease predisposition across
populations.
175
perform the genotyping in a cost-efficient way and, second, approaches which allow
the systematic conditional analysis on specific genetic variants using summary data,
without the need to share individualized genotypes, prohibited due to patient privacy
protections (Yang et al. 2012). Systematic application of these studies to hundreds of
genetic loci identified by GWAS promises to discover additional independent alleles
contributing to disease (explaining additional missing heritability) and a graded
structure of high- to low-risk haplotypes contributing to risk. These haplotypes will
then allow specific candidates for causality to emerge, which can then be tested in
functional or experimental systems.
7.4.4 Genetic Studies of Complex Traits across Ethnicities
Despite the higher prevalence and resulting disproportionate public health burden,
genetic studies of complex disease in non-European populations have not taken
off nearly at the rate of their European counterparts (Bustamante et al. 2011).
Fortunately, and recently, this trend appears to be shifting, for important scientific reasons. Well-powered GWAS for a range of disease traits across a range
of ethnic groups (e.g., Asian, Indian, African-American, Latino) have recently
been published (Saxena et al. 2012; SIGMA Type 2 Diabetes Consortium 2013;
DIAGRAM Consortium 2014) with many more underway. Many of these studies
are also actively working to pool data together and perform meta-analysis and
fine-mapping efforts (DIAGRAM Consortium 2014), forming collaborations with
existing groups who have focused on studies primarily in European populations.
These efforts are very important, as studies across populations worldwide can
identify new genetic risk factors that are more common or operate at stronger
effects across populations (Rosenberg et al. 2010; Pulit et al. 2010), and can
help improve the quantification of genetic risk prediction across ethnicities. A
potentially useful application of data across ethnicities is to exploit different patterns
of LD across groups to aid in fine-mapping efforts that seek to identify causal
alleles for disease. While this approach does have challenges, the combination
of the well-powered primary stages, custom-array genotyping technologies, and
a catalog of genetic variation down to low frequencies identified by the 1000
Genomes Project (1000 Genomes Project Consortium 2012) will test this hypothesis
very directly in the immediate future. Furthermore, at genomic sites where an
association has been established beyond a reasonable doubt, it is not unreasonable
to imagine that additional variants statistically independent of the established
association might exist and may segregate differently across ethnic groups, particularly if they are low-frequency or private to specific ethnicities. Thus, these
data may help not only to triangulate on the gene causal for disease but also
to help uncover the genetic mechanisms underlying disease predisposition across
populations.