Page 25 - The prevalence of the Val66Met polymorphism in musicians: Possible evidence for compensatory neuroplasticity from a pilot study

Lewis and Green Genome Medicine (2021) 13:14 Page 2 of 10

studies, most of the data available in the polygenic con- also captured. In addition to these forms of gene-
text to date has been from SNP-chips, which cheaply environment correlation, gene-by-environment interac-
probe a few hundred thousand, sometimes more than a tions, where the same variant has a different overall
million, of the more common variants within the effect in different environments, may also play a role. It
genome. is therefore not appropriate to straightforwardly draw
For most traits, currently available PRS fall far short of causal conclusions from PRS [18].
capturing the full variance of a trait as expected from It is against this context of issues with the basic sci-
heritability estimates. As larger data sets are studied, var- ence of PRS that their potential clinical application
iants with smaller effect sizes can be identified, and the needs to be assessed.
scores can capture more of the variance of the trait.
New statistical methodologies are also increasing the Are PRS ready for the clinic?
predictive power of these scores. A review of the con- In screening and in preventative medicine, the utility of
struction of PRS is given in Martin et al., where the au- specific tests and models is a complex arena [19]. Some
thors note the ease of producing PRS once the argue that a score has to give very high relative risks to
underlying data exists [9]. be useful as a screening measure, and PRS are nowhere
The scores themselves are normalized within a specific near this bar [20]. However, others have posited that
population and hence only give information about where PRS may already be clinically useful, and at the very
an individual falls within that population, for example least, we should be testing their use in the clinic [1, 21,
within the top 5%. To estimate relative and absolute 22]. Some large scale studies are already underway, for
risks, further work is needed. In addition to giving risk example the WISDOM trial is using PRS and other risk
information for developing a disease, some evidence sug- factors to stratify 100,000 women into higher and lower
gests that PRS can predict how likely someone is to re- risk for breast cancer [23], and a 20,000 person trial
spond to a treatment [10–12], for example through the eMERGE Network will return a number of
antidepressants [12, 13]. PRS for a variety of common conditions to each individ-
The underlying GWAS data lends itself to two inter- ual, and track the impact of this result disclosure [24].
linked problems in moving PRS into the clinic. First, The UK government is exploring a plan to return PRS
PRS suffer from a portability problem. A score based on to 5 million individuals through its Accelerating Detec-
individuals of one genetic ancestry can be multiple times tion of Disease Initiative [25].
less predictive in other ancestries [14]—the score does The potential benefits of identifying those at high risk of
not “port” well across populations. Some of this is be- developing a disease offer multiple approaches to pre-
cause of different patterns of linkage disequilibrium and ventative medicine [12]. Primary prevention, avoiding a
different allele frequencies in different populations. Most disease, could be achieved through lifestyle change or pro-
existing data are from individuals of European genetic active treatment. Secondary prevention, detecting a dis-
ancestry, so current scores are most predictive within ease early and preventing it from getting worse, could be
this population. Note that whereas race and ethnicity are achieved by more frequent screening of those at high risk
social constructs (either self-reported or assumed on the (e.g., mammograms for women at high risk of breast can-
basis of appearance, for example, by a healthcare pro- cer) or simply by helping increase uptake and compliance
vider), genetic ancestry is primarily a statistical concept with existing recommended screening [26]. Tertiary pre-
based on patterns of inheritance, though it nonetheless vention, trying to improve quality of life or reducing the
has ambiguities of interpretation [15]. Compounding this symptoms of an existing disease, could be achieved by bet-
portability issue, the scores are not always accurate even ter knowledge of treatment response, including responses
within individuals of the same genetic ancestry, but dif- to prescribed drugs. Somewhat more controversially, there
ferent demographics [16]. How the scores can be applied could be benefits in identifying those at lower risk, for ex-
across groups with different environmental exposures is ample in screening these individuals less often and hence
not well-characterized. saving resources and decreasing risks of over treatment
Second, PRS suffer from an interpretation problem:we [27]. With the introduction of any novel method to iden-
are not yet sure what conclusions can be drawn from tify those at high risk, the dangers of overdiagnosis (the
them. While some of the predictive signal captured in diagnosis of a condition that would never have caused
PRS comes from direct genetic effects, i.e., from variants symptoms or problems) and overtreatment (too much
that influence the phenotype of the individual, other treatment, in particular for insignificant disease) must be
predictive signals are also captured in a standard non- considered [28, 29].
family-based GWAS. These include indirect genetic Some data already exist for whether these benefits
effects, for example genetic nurture [17]. Effects of as- are realizable, but a clear picture is yet to emerge.
sortative mating and environmental confounding are Two retrospective analyses that integrated coronary