Efficient cross-trait penalized regression increases prediction accuracy in large cohorts using secondary phenotypes
Efficient cross-trait penalized regression increases prediction accuracy in large cohorts using secondary phenotypes
复制标题
DOI：
10.1038/s41467-019-08535-0
复制
发表时间：
2019-02-04
影响因子：
16.6
通讯作者：
Liang, Liming
中科院分区：
文献类型：
--
作者：
Chung, Wonil;Chen, Jun;Liang, Liming
我们介绍了跨性状惩罚回归（CTPR），这是一种强大而实用的多性状多基因风险预测方法。具体而言，我们提出了一种新的跨性状惩罚函数，结合Lasso和minimax凹惩罚（MCP），以结合大样本GWAS数据中多个性状之间的共享遗传效应。我们的方法从次要性状中提取信息，这些信息有利于基于个体水平基因型和/或汇总统计预测主要性状。我们对并行计算算法的新颖实现使得将我们的方法应用于生物库规模的GWAS数据成为可能。我们使用来自UK Biobank （N = 456,837）的大规模GWAS数据（类似于1M snp）来说明我们的方法。我们表明，我们的多性状方法在预测性能上优于最近提出的多性状GWAS分析（MTAG）。使用UK Biobank数据，BMI对身高的预测准确率从R-2 = 35.8% （MTAG）提高到42.5% （MCP + CTPR）或42.8% （Lasso + CTPR）
We introduce cross-trait penalized regression (CTPR), a powerful and practical approach for multi-trait polygenic risk prediction in large cohorts. Specifically, we propose a novel cross-trait penalty function with the Lasso and the minimax concave penalty (MCP) to incorporate the shared genetic effects across multiple traits for large-sample GWAS data. Our approach extracts information from the secondary traits that is beneficial for predicting the primary trait based on individual-level genotypes and/or summary statistics. Our novel implementation of a parallel computing algorithm makes it feasible to apply our method to biobank-scale GWAS data. We illustrate our method using large-scale GWAS data (similar to 1M SNPs) from the UK Biobank (N = 456,837). We show that our multi-trait method outperforms the recently proposed multi-trait analysis of GWAS (MTAG) for predictive performance. The prediction accuracy for height by the aid of BMI improves from R-2 = 35.8% (MTAG) to 42.5% (MCP + CTPR) or 42.8% (Lasso + CTPR) with UK Biobank data.