The False Dawn of Polygenic Risk Scores for Human Disease Prediction

Area of Science:

• Human genetics
• Genomic prediction
• Bioinformatics

Background:

• Polygenic risk scores (PRSs) leverage genome-wide association studies (GWAS) for disease risk prediction.
• Significant investments are made to enhance PRS accuracy through larger GWAS, advanced methods, and bias correction.

Purpose of the Study:

• To critically evaluate the assumptions underlying Polygenic risk scores (PRSs).
• To compare PRSs with estimated breeding values (EBVs) used in livestock selection.

Main Methods:

• Review of the historical development and underlying assumptions of PRSs.
• Comparative analysis of PRSs and EBVs in the context of genetic prediction and selection.

Main Results:

• The study questions the current credit given to PRSs by revisiting their foundational assumptions.
• A comparison with EBVs highlights potential limitations and areas for improvement in PRS methodology.

Conclusions:

• Re-evaluation of PRS assumptions is crucial, informed by the history of human genetics.
• Comparing PRSs to EBVs provides a novel perspective on genomic prediction accuracy and application.