Resampling-based tests for Lasso in genome-wide association studies.
Resampling-based tests for Lasso in genome-wide association studies.
复制标题
DOI：
10.1186/s12863-017-0533-3
复制
发表时间：
2017-07-24
期刊：
影响因子：
2.9
通讯作者：
Basu S
中科院分区：
文献类型：
--
作者：
Arbet J;McGue M;Chatterjee S;Basu S
全基因组关联研究涉及检测数百万遗传变异与性状之间的关联，通常使用单变量回归来测试每个单一变异与表型之间的关联。或者，Lasso惩罚回归允许人们联合建模所有遗传变异和表型之间的关系。然而，目前还不清楚如何最好地对各个Lasso系数进行推断，特别是在高维设置中。我们考虑六种方法来测试Lasso系数：两个排列（Lasso-Ayers，Lasso-PL）和一个分析方法（Lasso-AL）来选择惩罚参数的类型1-错误控制，残差自举（Lasso-RB），修改后的残差自举（Lasso-MRB），和排列测试（Lasso-PT）。方法进行了比较，通过模拟和应用明尼苏达中心的双胞胎和家庭研究。我们发现，对于有限的样本量与零预测，Lasso-RB，Lasso-MRB和Lasso-PT的数量不断增加，未能成为可行的推断方法。然而，Lasso-PL和Lasso-AL仍然是使用Lasso进行推理的快速而强大的工具，即使在高维中。我们的研究结果表明，拟议的排列选择程序（Lasso-PL）和分析选择方法（Lasso-AL）是快速和强大的替代标准单变量分析在全基因组关联研究。
Genome-wide association studies involve detecting association between millions of genetic variants and a trait, which typically use univariate regression to test association between each single variant and the phenotype. Alternatively, Lasso penalized regression allows one to jointly model the relationship between all genetic variants and the phenotype. However, it is unclear how to best conduct inference on the individual Lasso coefficients, especially in high-dimensional settings. We consider six methods for testing the Lasso coefficients: two permutation (Lasso-Ayers, Lasso-PL) and one analytic approach (Lasso-AL) to select the penalty parameter for type-1-error control, residual bootstrap (Lasso-RB), modified residual bootstrap (Lasso-MRB), and a permutation test (Lasso-PT). Methods are compared via simulations and application to the Minnesota Center for Twins and Family Study. We show that for finite sample sizes with increasing number of null predictors, Lasso-RB, Lasso-MRB, and Lasso-PT fail to be viable methods of inference. However, Lasso-PL and Lasso-AL remain fast and powerful tools for conducting inference with the Lasso, even in high-dimensions. Our results suggest that the proposed permutation selection procedure (Lasso-PL) and the analytic selection method (Lasso-AL) are fast and powerful alternatives to the standard univariate analysis in genome-wide association studies.
登录
查看更多内容
影响因子：
2.6
作者：
Hicks, Brian M.;Schalet, Benjamin D.;Malone, Stephen M.;Iacono, William G.;McGue, Matt
通讯作者：
McGue, Matt
影响因子：
0.9
作者：
Miller, Michael B.;Basu, Saonli;Cunningham, Julie;Eskin, Eleazar;Malone, Steven M.;Oetting, William S.;Schork, Nicholas;Sul, Jae Hoon;Iacono, William G.;McGue, Matt
通讯作者：
McGue, Matt
影响因子：
4.3
作者：
Hoffman GE;Logsdon BA;Mezey JG
通讯作者：
Mezey JG
影响因子：
5.4
作者：
HILL W G;ROBERTSON A
通讯作者：
ROBERTSON A
DOI：
10.1080/00949659908811936
发表时间：
1999-01-01
影响因子：
1.2
作者：
Anderson, MJ;Legendre, P
通讯作者：
Legendre, P