Polygenic Risk Scores: Why They Matter and How We Can Make Them Work for Everyone

Polygenic risk scores (PRSs) are becoming increasingly important in modern medicine. They represent an individual's genetic predisposition to a particular trait or disease, calculated by summing the effects of many genetic variants. PRSs have the potential to revolutionize healthcare by improving disease prediction, enabling earlier diagnoses, andPersonalized treatment strategies.

However, a significant challenge exists: PRSs developed primarily from data of European ancestry often perform poorly in individuals from diverse genetic backgrounds. This disparity could worsen existing health inequities. To address this, scientists are working to understand the factors limiting PRS transferability and develop methods to improve their accuracy across all populations.

Factors Affecting PRS Accuracy
Several factors influence how well a PRS performs across different populations:

* Genetic Factors:

* Heritability: The proportion of trait variation attributed to genetics can differ across populations, affecting PRS accuracy.

* Allele Frequencies and Effect Sizes: Differences in the frequencies of causal variants and their effects can impact PRS transferability.

* Linkage Disequilibrium (LD): Variations in LD patterns across populations can also affect PRS performance. For instance, African populations generally have smaller LD blocks compared to European or East Asian populations, requiring a larger number of variants to capture the same amount of genetic variation.

* Social and Environmental Factors: Environmental exposures, lifestyle choices, and social determinants of health (SDOH) can all influence disease risk and PRS performance. SDOH include factors like access to healthcare, socioeconomic status, and exposure to racism and discrimination. These factors can act as effect modifiers, confounders, or mediators of PRS effects.

Improving PRS Transferability
To make PRSs more equitable, researchers are developing new methods that integrate data from diverse populations. These methods generally fall into a few categories:

* Combining Population-Specific Summary Statistics: This involves combining genome-wide association studies (GWAS) data from multiple ancestry groups using meta-analysis. This approach assumes that the genetic effects are the same across all ancestries, which is not always the case.

* Joint Modeling of Multiple Populations: This approach uses a large-scale GWAS, often from European populations, to improve prediction in non-European populations with smaller GWAS. These methods leverage the idea that many causal genetic signals are shared across populations.

* Incorporating Additional Information: Some methods incorporate functional annotations or data from genetically correlated traits to improve PRS construction.

Considerations for Diverse Populations
Despite these advances, several challenges remain:

* Validation Data: Many methods require validation datasets with individual-level data to fine-tune their algorithms. However, such datasets may not be readily available in diverse non-European populations.

* Categorizing Genetic Ancestry: Many methods require assigning individuals to specific genetic ancestry groups, which can be challenging in admixed populations.

* Computational Complexity: Balancing prediction accuracy with computational efficiency is an ongoing challenge. More complex models can be difficult to implement on a genome-wide scale.

Evaluating Clinical Utility
Demonstrating the clinical utility of PRSs across diverse populations requires careful selection of performance metrics. While the area under the receiver operating characteristic curve (AUC) is commonly used, it doesn't account for disease rates in different populations and may not provide insight into clinical utility. Metrics such as absolute risk, positive and negative predictive values, and net reclassification indices are more clinically relevant.

Future Directions
The future of PRS research depends on several key steps:

* Increase Diversity in Genomic Studies: Expanding genomic studies to include underrepresented populations is essential for developing more accurate and generalizable PRSs.

* Improve Data Collection and Harmonization: Standardizing phenotyping and data collection methods across diverse populations will facilitate data aggregation and pooling efforts.

* Address Social Determinants of Health: Understanding and accounting for the complex interplay between genetic ancestry, SDOH, and environmental factors is crucial for contextualizing PRS performance and informing public health interventions.

* Develop Universal PRSs: Creating PRSs that can be applied across diverse populations while considering fairness constraints will help ensure equitable performance.

* Contextualize PRS Performance in Healthcare Settings: Integrating PRSs with existing clinical predictors and risk assessment tools will facilitate their adoption in real-world healthcare settings.

By addressing these challenges and continuing to innovate in PRS methodology, data generation, and clinical implementation, the promise of precision medicine can be realized for all populations.

Disclaimer: This blog post is based on the provided research article and is intended for informational purposes only. It is not intended to provide medical advice. Please consult with a healthcare professional for any health concerns.

References:
Kachuri, L., Chatterjee, N., Hirbo, J. et al. Principles and methods for transferring polygenic risk scores across global populations. Nat Rev Genet 25, 8–25 (2024). https://doi.org/10.1038/s41576-023-00637-2