Polygenic Risk Scores

A polygenic risk score (PRS) is a numerical estimate of a person's genetic susceptibility to a disease or trait, based on many genetic variants across the genome. It combines the risk contributions of these variants, usually using effect estimates obtained from genome-wide association studies (GWAS). It indicates relative genetic risk within a reference population, not a certain diagnosis or complete prediction of disease.

How it is constructed

Researchers identify variants statistically associated with a trait in large genomic datasets. For an individual, the number of risk-associated alleles at each selected variant is multiplied by its estimated effect size, and these weighted contributions are added to produce the polygenic score.

• A PRS captures the combined influence of many variants, each of which may individually have only a small effect.
• The score is commonly interpreted by comparing a person with the risk distribution in a reference population.

Uses in genomic medicine

A PRS can support risk stratification by identifying groups with relatively higher or lower inherited susceptibility. When validated, it may complement family history, clinical measurements and environmental risk factors.

• It can inform research on disease susceptibility and population-level prevention.
• It may help tailor the timing or intensity of screening, but should not be used as a stand-alone clinical decision.

Limitations and database diversity

A PRS is probabilistic because disease also depends on environment, lifestyle, age and other biological factors. Its accuracy depends heavily on the size, quality and population composition of the genomic databases used to develop and validate it.

• Scores developed mainly from one ancestry group may transfer poorly to another because allele frequencies and patterns of linkage disequilibrium differ across populations.
• Under-representation of populations can widen health inequities if less accurate scores are applied without suitable validation.
• Privacy, informed consent, possible discrimination and responsible communication are important ethical concerns.