PXS

PXS integrates genetic and non-genetic factors to compute a poly-exposure score for predicting individual risk of common diseases.

Key Features:

• UK Biobank data: Developed through analysis of data from 459,613 participants in the UK Biobank.
• Integration of non-genetic factors: Includes 96 environmental exposures such as lifestyle choices, employment status, diet, and pollution levels.
• Demographic considerations: Accounts for demographic variables including sex and age.
• Additive modelling approach: Employs an additive modelling technique to estimate and validate the poly-exposure score.
• Application across multiple diseases: Validated for atrial fibrillation, coronary artery disease, inflammatory bowel disease, and type 2 diabetes.
• Superior predictive performance: In comparative analyses against polygenic risk scores (PRSs), PXS achieved a maximum AUC of 0.828 for type 2 diabetes versus 0.711 for PRSs.
• Identification of high-risk individuals: Detects individuals with low genetic risk but high overall disease risk due to non-genetic factors.

Scientific Applications:

• Clinical risk prediction: Provides combined genetic and exposure-informed risk estimates to support individualized risk stratification.
• Public health strategies: Informs population-level interventions by quantifying the impact of modifiable environmental exposures on disease risk.
• Research and development: Enables investigation of the combined effects of genetic and non-genetic factors on disease etiology and risk prediction performance.

Methodology:

Analysis of UK Biobank data from 459,613 participants using an additive modelling framework to integrate genetic variables, 96 environmental exposures and demographic covariates, with comparative performance evaluation via AUC against PRSs.