利用基于汇总统计量的选择和收缩（S4）和 LDpred2 提高复杂性状的多基因评分。
Improving on polygenic scores across complex traits using select and shrink with summary statistics (S4) and LDpred2.
机构信息
Centre for Cancer Genetic Epidemiology, Department of Public Health and Primary Care, University of Cambridge, Cambridge, CB1 8RN, UK.
Department of Computational Biomedicine, Cedars-Sinai Medical Center, California, 90048, United States of America.
出版信息
BMC Genomics. 2024 Sep 18;25(1):878. doi: 10.1186/s12864-024-10706-3.
BACKGROUND
As precision medicine advances, polygenic scores (PGS) have become increasingly important for clinical risk assessment. Many methods have been developed to create polygenic models with increased accuracy for risk prediction. Our select and shrink with summary statistics (S4) PGS method has previously been shown to accurately predict the polygenic risk of epithelial ovarian cancer. Here, we applied S4 PGS to 12 phenotypes for UK Biobank participants, and compared it with the LDpred2 and a combined S4 + LDpred2 method.
RESULTS
The S4 + LDpred2 method provided overall improved PGS accuracy across a variety of phenotypes for UK Biobank participants. Additionally, the S4 + LDpred2 method had the best estimated PGS accuracy in Finnish and Japanese populations. We also addressed the challenge of limited genotype level data by developing the PGS models using only GWAS summary statistics.
CONCLUSIONS
Taken together, the S4 + LDpred2 method represents an improvement in overall PGS accuracy across multiple phenotypes and populations.
背景
随着精准医学的发展，多基因评分（PGS）在临床风险评估中变得越来越重要。已经开发出许多方法来创建具有更高准确性的多基因模型，以进行风险预测。我们之前使用的选择和缩减与汇总统计信息（S4）PGS 方法已经被证明可以准确预测上皮性卵巢癌的多基因风险。在这里，我们将 S4 PGS 应用于 UK Biobank 参与者的 12 种表型，并将其与 LDpred2 和组合的 S4 + LDpred2 方法进行了比较。
结果
S4 + LDpred2 方法在 UK Biobank 参与者的各种表型中提供了整体提高的 PGS 准确性。此外，S4 + LDpred2 方法在芬兰和日本人群中具有最佳的估计 PGS 准确性。我们还通过仅使用 GWAS 汇总统计信息来开发 PGS 模型，解决了基因型水平数据有限的挑战。
结论
总之，S4 + LDpred2 方法在多个表型和人群中代表了整体 PGS 准确性的提高。
Zotero 插件