Making Polygenic Risk Scores Work for Everyone
The advent of polygenic risk scores (PRSs) has revolutionized our understanding of genetic predispositions to complex diseases. These scores, derived from genome-wide association studies (GWAS), aggregate the influence of multiple genetic variants to predict an individual’s risk for developing specific conditions. However, a critical challenge persists: how to ensure PRSs are predictive and equitable across diverse global populations. The recent review by Kachuri et al. in Nature Reviews Genetics delves into the nuances of transferring PRSs to diverse ancestries, shedding light on methodologies, challenges, and the path forward.
The Promise and Peril of PRS
Polygenic risk scores distill complex genetic interactions into actionable insights, potentially enhancing precision medicine. Yet, as over 85% of GWAS participants hail from European ancestries, the predictive power of PRSs for non-European populations often falls short. This disparity risks exacerbating health inequities, particularly as PRS applications expand into clinical settings.
Genetic and Non-Genetic Barriers to PRS Transferability
The transferability of PRSs across populations is hindered by both genetic and environmental factors. Key genetic barriers include:
Causal Variant Frequency: Variants common in one population may be rare in another, limiting PRS accuracy.
Allelic Effect Sizes: Differences in the magnitude of genetic effects across ancestries further complicate predictions.
Linkage Disequilibrium (LD) Patterns: Variations in LD—the correlation of genetic variants—affect how well causal variants are tagged by GWAS.
Additionally, social determinants of health (SDOH), such as socioeconomic status, discrimination, and environmental exposures, play critical roles. These non-genetic factors often intersect with genetic ancestry, creating confounding influences that impact PRS utility.
Innovative Methods to Bridge the Gap
To improve PRS transferability, researchers have devised several strategies:
Multi-Ancestry GWAS: Combining data from diverse populations enhances the representation of global genetic diversity.
Admixture-Aware Models: Methods such as local ancestry inference allow for finer resolution in admixed populations.
Cross-Ancestry Genetic Correlations: Leveraging shared genetic architectures across populations aids in refining PRS models.
Emerging tools like PRS-CSx and TL-PRS integrate GWAS summary statistics from multiple populations, demonstrating promise for enhancing prediction accuracy.
Evaluating Clinical Utility
The clinical applicability of PRSs depends on robust evaluation metrics. While measures like the area under the curve (AUC) assess discriminatory power, they do not address calibration across populations. Cross-population calibration and integration with absolute risk models are vital for equitable clinical deployment.
Future Directions
The road ahead requires a multi-pronged approach:
Diverse Data Collection: Expanding GWAS to underrepresented populations is paramount. Initiatives like the All of Us Research Program aim to close this gap.
Ethical Considerations: Transparent data-sharing practices and community engagement are essential to build trust and ensure equitable benefits.
Integration with SDOH: Incorporating social and environmental factors into PRS models will enhance their real-world applicability.
Conclusion
Kachuri et al. highlight the transformative potential of polygenic risk scores in precision medicine while emphasizing the need for inclusivity. By addressing both genetic and non-genetic barriers, we can move closer to a future where PRSs serve as tools for global health equity.
References:
Kachuri, L., et al. (2024). Principles and methods for transferring polygenic risk scores across global populations. Nature Reviews Genetics, 25, 8–25.
