>_ Why This Matters: From single SNPs to genome-wide scores
Previous analyses established that many common health traits are highly polygenic.
If thousands of variants each have tiny effects, looking at one SNP at a time becomes inefficient and potentially misleading. Polygenic risk scores attempt to solve that problem.
Instead of reading thousands of variants independently, researchers combine their estimated statistical effects into one single score.
Compressing thousands of genetic observations into a single number creates a new challenge: how should that number be interpreted?
>_ Percentile Is Not Probability
This distinction is critical.
Suppose your PRS is in the 95th percentile. That means your score is higher than approximately 95% of people in the reference population used for comparison.
It does NOT mean you have a 95% probability of developing the condition.
Percentile rank and absolute disease risk are completely different quantities. Converting a score into personal probability requires combining it with baseline disease incidence, age, and clinical risk factors.
>_ PRS Is Not the Same as a Pathogenic Monogenic Mutation
A rare pathogenic variant in a disease-associated gene (like BRCA1) can confer a large effect through a direct biological mechanism.
A polygenic score aggregates hundreds or thousands of common variants with small individual effects.
Both are genetic information, but they represent fundamentally different architectures. A high PRS should not automatically be described as equivalent to “having a mutation.”
>_ The Ancestry Portability Problem
Historically, many large GWAS datasets disproportionately represented individuals of European ancestry.
Because allele frequencies and patterns of linkage disequilibrium differ between populations, a PRS developed in one ancestry group can lose predictive accuracy when applied to another.
Recent clinical implementation efforts (such as eMERGE 2024 and ICDA guidelines) focus heavily on multi-ancestry validation and proper score calibration.
>_ What Else Does a PRS Leave Out?
Diet, sleep, physical activity, smoking, pollution, and social conditions do not disappear because a PRS exists.
Blood pressure, lipids, and blood glucose contain direct live information about your current phenotype.
Family history captures shared genetic and non-genetic factors that a common-variant PRS may miss.
Common-variant PRSs generally do not capture rare, high-impact monogenic mutations.