Polygenic Risk Scores That Predict Common Diseases Using Millions of Single Nucleotide Polymorphisms: Is More, Better?

In recent studies, this increased risk is claimed to be similar to conditions associated with known causal genetic mutations. [...]the argument goes, it is time to start thinking about implementing PRSs in routine clinical care (1, 2). [...]another study (5) also investigated polygenic risk prediction of CAD using 1.7 million SNPs in the UK Biobank and found a similar high AUC (0.79) for the prediction ofprevalent cases that included age but a low AUC (0.62) for the prediction of incident cases in which age was used as the time variable. [...]the authors concluded that people at high polygenic risk would not have been identified using clinical risk factors because they had similar means and frequencies of clinical risk factors to the rest. Moving Forward The increasing interest in the study of PRSs warrants a reflection on what evidence is needed to move the scores to healthcare practice, if proven effective. Because the predictive ability of risk models is known to vary between populations, it is crucial that a prediction analysis be conducted in a population that is representative for the population in which the PRS is intended to be applied: