IPRS: Leveraging Gene-Environment Interaction to Reconstruct Polygenic Risk Score.

Area of Science:

• Genetics
• Epidemiology
• Biostatistics

Background:

• Polygenic risk scores (PRS) predict genetic susceptibility but traditional models overlook gene-environment (GxE) interactions.
• This limitation hinders accurate prediction and risk stratification for complex diseases.

Purpose of the Study:

• To develop and evaluate an interaction polygenic risk score (iPRS) model that incorporates GxE interactions.
• To compare the predictive performance of iPRS against traditional PRS.

Main Methods:

• Developed the iPRS method to reconstruct PRS by leveraging GxE interactions.
• Evaluated iPRS using simulations and real-world data from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial.
• Assessed prediction performance, risk stratification, and calibration.

Main Results:

• Simulations showed iPRS improved disease risk prediction, risk stratification, and calibration compared to traditional PRS.
• iPRS demonstrated superior predictive effect in the PLCO lung cancer data analysis (p = 0.0205).
• Performance gains were particularly notable when antagonistic GxE interactions were prevalent.

Conclusions:

• The iPRS model shows significant potential for predicting disease risk and optimizing high-risk population screening.
• iPRS can enhance the clinical benefits of preventive interventions by providing more accurate risk assessments.