Leveling the Playing Field: New Trans-Ancestry Polygenic Risk Score Transforms Type 1 Diabetes Screening

Highlights

• Development of a novel trans-ancestry polygenic risk score (TA-PS) for type 1 diabetes that outperforms previous European-centric models.
• The TA-PS achieved a significantly higher AUROC (0.89) compared to the standard GRS2x (0.85) in multi-ancestry populations.
• Enhanced sensitivity in non-European populations, including South Asians and African Americans, supporting more equitable disease screening.
• Validation across multiple independent cohorts, including the UK Biobank and All of Us, confirms the tool’s robustness for population-wide surveillance.

The Challenge of Genetic Diversity in Type 1 Diabetes

Type 1 diabetes (T1D) is a chronic autoimmune condition characterized by the destruction of insulin-producing beta cells. It is one of the most heritable common diseases, with genetics playing a pivotal role in determining susceptibility. Over the last decade, polygenic risk scores (PRSs) have emerged as powerful tools to identify individuals at high risk for T1D before the onset of clinical symptoms. Early identification is critical; it allows for autoantibody surveillance and metabolic monitoring, which can prevent life-threatening diabetic ketoacidosis (DKA) at diagnosis and enable early access to emerging preventive therapies, such as teplizumab.

However, a significant clinical and ethical hurdle has persisted: most PRSs for T1D were developed using data primarily from individuals of European ancestry. When applied to non-European populations, these scores often exhibit a substantial drop in predictive accuracy. This “genomic gap” risks exacerbating health disparities, as screening programs based on European data may fail to identify high-risk individuals in minority or multi-ethnic communities. To address this, Jumentier et al. (2026) developed and validated a trans-ancestry polygenic risk score (TA-PS) designed for equitable performance across diverse global populations.

Study Design: Building a More Inclusive Risk Engine

The research team utilized the PRS-CSx method, a sophisticated Bayesian framework that integrates summary statistics from multiple ancestry-specific genome-wide association studies (GWAS) to improve cross-population prediction. The development phase utilized data from large European, East Asian, African American, and Hispanic T1D cohorts, totaling 29,469 cases.

The Architecture of TA-PS

The TA-PS consists of two primary components:

• A non-HLA component: This incorporates over one million genetic variants across the genome, optimized for trans-ancestry performance.
• An HLA component: This utilizes the Human Leukocyte Antigen (HLA) component from the established European GRS2x, which remains the strongest genetic predictor of T1D.

Validation Cohorts

The performance of the TA-PS was first tested in a multi-ancestry cohort from Montreal, Canada (N=4,657). To ensure the findings were generalizable, the researchers validated the score in four independent datasets: the Children’s Hospital of Philadelphia (CHOP) Center for Applied Genomics, the Genetic Risk Assessment for Chinese Eaglet-T1D (GRACE) cohort, the All of Us Research Program, and the UK Biobank.

Key Findings: Elevating Performance Across Ancestries

The results of the study represent a significant advancement in precision medicine for diabetes. In the primary Montreal-based multi-ancestry cohort, the TA-PS demonstrated an Area Under the Receiver Operating Curve (AUROC) of 0.89. This was a statistically significant improvement over the European-based GRS2x, which yielded an AUROC of 0.85 in the same population.

Sensitivity and Specificity Improvements

One of the most clinically relevant findings was the improvement in sensitivity—the ability of the test to correctly identify those who will develop the disease. Using a 90th percentile cut-off, the TA-PS achieved a sensitivity of 0.71 in Europeans and 0.77 in South Asians. In contrast, the GRS2x had previously shown a sensitivity as low as 0.32 in African American populations and 0.56 in Europeans within diverse contexts. While the specificity of the TA-PS was slightly lower than that of the GRS2x, it remained well within the acceptable clinical range (≥0.83 across all ancestries).

Global Validation

In the Chinese GRACE cohort, the TA-PS maintained high predictive power, confirming its utility in East Asian populations. Similarly, validation in the All of Us and UK Biobank cohorts—which include broad demographic representations—consistently showed that the TA-PS provided more balanced and accurate risk assessments than previous models. This consistency across cohorts suggests that the TA-PS is ready for integration into large-scale, population-wide screening programs.

Expert Commentary: Clinical Utility and Future Directions

The development of the TA-PS is a landmark achievement in addressing the “polygenic risk gap.” For clinicians, the ability to accurately assess risk in non-European patients is not just a matter of scientific accuracy but of health equity. As we enter the era of disease-modifying therapies for T1D, having a screening tool that works for everyone is essential.

Mechanistic Insights

The success of the TA-PS likely stems from its ability to capture both universal risk variants and those that are specific to or more common in certain ancestral groups. By combining a dense non-HLA polygenic component with the highly impactful HLA region, the model captures a more complete picture of the genetic architecture of T1D. This suggests that while the HLA region is the dominant driver of risk globally, the “fine-tuning” provided by a million other variants is what allows for high-resolution risk stratification in diverse populations.

Limitations and Considerations

While the TA-PS is a major step forward, researchers note that genetic risk is only one piece of the puzzle. Environmental factors, which may also vary by ancestry and geography, play a significant role in triggering the autoimmune process. Additionally, while the TA-PS performs better than its predecessors, ongoing efforts to collect even larger GWAS datasets from African and indigenous populations will be necessary to further refine these scores.

Conclusion: A Milestone for Global Public Health

The TA-PS represents a significant evolution in T1D risk prediction. By outperforming the European-based GRS2x and providing comparable accuracy across different ancestries, it removes a major barrier to the implementation of universal screening. This tool holds the potential to reduce the incidence of DKA and ensure that all children, regardless of their ethnic background, have the opportunity to benefit from early intervention and preventive care.

References

Jumentier B, Qu HQ, Lu T, et al. Development and validation of a trans-ancestry polygenic risk score for type 1 diabetes. Diabetologia. 2026. PMID: 41795035.

Khera AV, Chaffin M, Aragam KG, et al. Genome-wide polygenic scores for common diseases identify individuals with risk equivalent to monogenic mutations. Nat Genet. 2018;50(9):1219-1224.

Sharp SA, Rich SS, Wood AR, et al. Development and Standardization of an Improved Type 1 Diabetes Genetic Risk Score for Use in Newborn Screening and Incident Diagnosis. Diabetes Care. 2019;42(2):200-207.