Lee Phil H, Jung Jae-Yoon, Sanzo Brandon T, Duan Rui, Ge Tian, Waldman Irwin, Smoller Jordan W, Schwaba Ted, Tucker-Drob Elliot M, Grotzinger Andrew D
Psychiatric and Neurodevelopmental Genetics Unit, Center for Genomic Medicine, Mass General Brigham, Boston, MA, USA.
Department of Psychiatry, Mass General Brigham and Harvard Medical School, Boston, MA, USA.
medRxiv. 2025 Mar 28:2025.03.26.25324720. doi: 10.1101/2025.03.26.25324720.
Psychiatric disorders exhibit substantial genetic overlap, raising questions about the utility of transdiagnostic genetic risk models. Using data from the Research Program (N=102,091), we evaluated common psychiatric genetic (CPG) factor-based polygenic risk scores (PRSs) compared to standard disorder-specific PRSs. The CPG PRS consistently outperformed disorder-specific scores in predicting individual disorder risk, explaining 1.07 to 24.6 times more phenotypic variance across 11 psychiatric conditions. Meanwhile, many disorder-specific PRSs retained independent but smaller contributions, highlighting the complementary nature of shared and disorder-specific genetic risk. While alternative multi-factor models improved model fit, the CPG PRS provided comparable or superior predictive performance across most disorders, including overall comorbidity burden. Cross-ancestry analyses however revealed notable limitations of European-centric GWAS datasets for other populations due to ancestral differences in genetic architecture. These findings underscore the potential value of transdiagnostic PRSs for psychiatric genetics while highlighting the need for more equitable genetic risk models.
精神疾病存在大量的遗传重叠，这引发了关于跨诊断遗传风险模型效用的问题。利用来自研究项目（N = 102,091）的数据，我们评估了基于常见精神疾病遗传（CPG）因素的多基因风险评分（PRSs），并与标准的特定疾病PRSs进行了比较。在预测个体疾病风险方面，CPG PRS始终优于特定疾病评分，在11种精神疾病中解释的表型变异比特定疾病评分多1.07至24.6倍。同时，许多特定疾病的PRSs保留了独立但较小的贡献，凸显了共享遗传风险和特定疾病遗传风险的互补性质。虽然替代多因素模型改善了模型拟合，但CPG PRS在大多数疾病中提供了相当或更好的预测性能，包括总体共病负担。然而，跨祖先分析揭示了以欧洲为中心的全基因组关联研究（GWAS）数据集由于遗传结构的祖先差异而对其他人群存在显著局限性。这些发现强调了跨诊断PRSs在精神疾病遗传学中的潜在价值，同时突出了需要更公平的遗传风险模型。
