Answers are grounded in this paper and cite their sources.
Early breast cancer detection is crucial for positive patient outcomes, with 5-year survival rates exceeding 90% when diagnosed early. Genome-wide association studies and polygenic risk scores (PRSs) have emerged as valuable tools for risk stratification, particularly given breast cancer’s substantial incidence and heritability. This study analyzed 12,480 subjects from China Medical University Hospital using published PRSs from the PRS Catalog to evaluate breast cancer risk prediction in the Taiwanese population. Our comprehensive analysis incorporated 64 previously described PRSs to compare their predictive performance and distributions within the Taiwanese population. To identify the optimal PRS for the Taiwanese population, we developed a meta-score (metaPRS) based on genetic algorithm to select highly impactful variants from all the included PRS. Notably, several genetic variants showed consistent significance across multiple PRS models, with rs11249433, rs10941679, and rs10816625 appearing in more than 30 different PRSs. The performance evaluation of the metaPRS revealed a notable distinction between cases and controls, with an odds ratio of 14.6 and 0.691 of area under the curve, which is higher than most published PRSs. Also, metaPRS demonstrated strong generalizability in the independent validation cohort, yielding an odds ratio of 1.46 and an AUROC of 0.542. These findings highlight the potential utility of PRSs in the Taiwanese population while also demonstrating the importance of population-specific considerations in PRS applications. This study provides valuable insights through comparative analysis of diverse PRSs in the Taiwanese population, contributing to our understanding of how published PRSs performs across different ethnic groups and informing future developments in genetic risk prediction.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: