Improving polygenic risk prediction of renal function by removing biomarker-specific effects

Area of Science:

• Genetics and Genomics
• Clinical Biomarkers
• Renal Physiology

Background:

• Biomarkers are vital in clinical practice, but genetic factors can independently alter their baseline levels, complicating result interpretation.
• Estimated glomerular filtration rate (eGFR) equations, commonly used in clinical settings, rely on biomarkers like cystatin C and creatinine.
• Discrepancies in eGFR estimates may arise from the underlying genetic architecture of these biomarkers.

Purpose of the Study:

• To investigate how the genetics of biomarkers impact measurement variability.
• To determine if genetic factors explain discrepancies among different eGFR equations.
• To identify shared and biomarker-specific genetic regions influencing renal function biomarkers.

Main Methods:

• Utilized shared genetic architecture to identify "shared" (renal-specific) and "biomarker-specific" genomic regions.
• Developed polygenic risk scores (PRS) based on these identified genetic regions.
• Assessed the explanatory power of different PRS for variability in creatinine-derived eGFR estimates.

Main Results:

• A shared, renal-specific polygenic risk score (PRS) explained 60% more variability in creatinine-derived eGFR estimates compared to biomarker-specific PRS.
• Biomarker-specific genetic regions also contributed to measurement variability.
• Integrating all genetic regions in a PRS did not outperform the shared PRS in explaining eGFR variability.

Conclusions:

• Biomarker-specific genetics are essential for constructing accurate polygenic risk scores for eGFR.
• Genetic variations significantly impact biomarker levels and the interpretation of clinical risk estimates.
• Future PRS development for eGFR and other biomarker-derived estimates must account for underlying genetic influences.