7 What Have We Learned from GWAS?
169
7.3.4 Complex Traits Share a Genetic Basis in Common Across
Populations
One observation from GWAS is that the genetic basis of this disease is indeed
shared across populations descended from different ancestries, though the details
can vary across associated sites. Taking T2D as an example trait, this has been
primarily observed, where sufficient statistical power exists, at genetic associations
identified in one population that also associate (individually or in aggregate) in
other populations (Waters et al. 2010; Saxena et al. 2012). In several cases, the
same physical regions and even the same SNPs have been implicated by association
studies across ethnic groups (e.g., TCF7L2 or KCNQ1), which imply the existence
of a common underlying etiology and mechanism. However, in other cases, different
or additional genetic factors at the locus are also associated given the best available
data (e.g., HMGA2 or CDKN2A/2B), and in some cases, the estimated effect of
genotype to phenotype is heterogeneous across populations (e.g., CDKAL1). These
observations could be compatible with shared biology, but variable mechanisms
underlie disease susceptibilities across populations. Alternatively, the data could
be compatible with simply a lack of statistical power in testing due to small
sample sizes coupled with a lack of a comprehensive variant map across ethnicities
to resolve the true, underlying causal variant(s) associated across groups. While
a comprehensive answer to this question will require extensive fine mapping
and genotyping in large samples across ethnicities, it is overwhelmingly clear
that further genetic investigations along this track are essential to identify novel
biological underpinnings of disease.
7.3.5 The Identification of Unknown Mechanisms for Disease
Pathogenesis
The preoccupation with reconciling the question of missing heritability is certainly
important for many applications, but should not detract from the goal of translating
genetic findings into biological knowledge, hypotheses for a mechanism, and
actionable intelligence for treatment. The most promising aspect of GWAS is that
the design has provided a rigorous procedure by which biology and mechanisms
related to disease can be discovered. This can clearly be seen as, for a range of
cardiovascular, anthropometric, and glycemic traits, previously known monogenic
causes of these diseases are preferentially found by common variant association
(Voight et al. 2010; Teslovich et al. 2010). But further still, it is also increasingly clear that previously unknown causes and mechanisms of disease have also
been clearly identified by GWAS. Those examples include, but are not limited
to, the complement pathways for age-related macular degeneration, autophagy,
Th17, and interleukin-23 pathways in Crohn’s disease, multiple sclerosis, and
other autoimmune-mediated diseases, hedgehog and TGF-β signaling, chromatin
remodeling, histones for human stature (Lango Allen et al. 2010), and gene
169
7.3.4 Complex Traits Share a Genetic Basis in Common Across
Populations
One observation from GWAS is that the genetic basis of this disease is indeed
shared across populations descended from different ancestries, though the details
can vary across associated sites. Taking T2D as an example trait, this has been
primarily observed, where sufficient statistical power exists, at genetic associations
identified in one population that also associate (individually or in aggregate) in
other populations (Waters et al. 2010; Saxena et al. 2012). In several cases, the
same physical regions and even the same SNPs have been implicated by association
studies across ethnic groups (e.g., TCF7L2 or KCNQ1), which imply the existence
of a common underlying etiology and mechanism. However, in other cases, different
or additional genetic factors at the locus are also associated given the best available
data (e.g., HMGA2 or CDKN2A/2B), and in some cases, the estimated effect of
genotype to phenotype is heterogeneous across populations (e.g., CDKAL1). These
observations could be compatible with shared biology, but variable mechanisms
underlie disease susceptibilities across populations. Alternatively, the data could
be compatible with simply a lack of statistical power in testing due to small
sample sizes coupled with a lack of a comprehensive variant map across ethnicities
to resolve the true, underlying causal variant(s) associated across groups. While
a comprehensive answer to this question will require extensive fine mapping
and genotyping in large samples across ethnicities, it is overwhelmingly clear
that further genetic investigations along this track are essential to identify novel
biological underpinnings of disease.
7.3.5 The Identification of Unknown Mechanisms for Disease
Pathogenesis
The preoccupation with reconciling the question of missing heritability is certainly
important for many applications, but should not detract from the goal of translating
genetic findings into biological knowledge, hypotheses for a mechanism, and
actionable intelligence for treatment. The most promising aspect of GWAS is that
the design has provided a rigorous procedure by which biology and mechanisms
related to disease can be discovered. This can clearly be seen as, for a range of
cardiovascular, anthropometric, and glycemic traits, previously known monogenic
causes of these diseases are preferentially found by common variant association
(Voight et al. 2010; Teslovich et al. 2010). But further still, it is also increasingly clear that previously unknown causes and mechanisms of disease have also
been clearly identified by GWAS. Those examples include, but are not limited
to, the complement pathways for age-related macular degeneration, autophagy,
Th17, and interleukin-23 pathways in Crohn’s disease, multiple sclerosis, and
other autoimmune-mediated diseases, hedgehog and TGF-β signaling, chromatin
remodeling, histones for human stature (Lango Allen et al. 2010), and gene