Shrinkage-based diagonal Hotelling's tests for high-dimensional small sample size data

JOURNAL OF MULTIVARIATE ANALYSIS

卷:143

Article

Dong, Kai¹  Pang, Herbert^2,3  Tong, Tiejun¹  Genton, Marc G.⁴ 

[1] Hong Kong Baptist Univ, Dept Math, Hong Kong, Hong Kong, Peoples R China

[2] Univ Hong Kong, Sch Publ Hlth, Hong Kong, Hong Kong, Peoples R China

[3] Duke Univ, Dept Biostat & Bioinformat, Durham, NC 27706 USA

[4] King Abdullah Univ Sci & Technol, CEMSE Div, Thuwal, Saudi Arabia

关键词: Diagonal Hotelling's test;    High-dimensional data;    Microarray data;    Null distribution;    Optimal variance estimation;   

DOI  :  10.1016/j.jmva.2015.08.022

来源: Elsevier

【 摘 要 】

High-throughput expression profiling techniques bring novel tools and also statistical challenges to genetic research. In addition to detecting differentially expressed genes, testing the significance of gene sets or pathway analysis has been recognized as an equally important problem. Owing to the large p small n paradigm, the traditional Hotelling's T-2 test suffers from the singularity problem and therefore is not valid in this setting. In this paper, we propose a shrinkage-based diagonal Hotelling's test for both one-sample and two-sample cases. We also suggest several different ways to derive the approximate null distribution under different scenarios of p and n for our proposed shrinkage-based test. Simulation studies show that the proposed method performs comparably to existing competitors when n is moderate or large, but it is better when n is small. In addition, we analyze four gene expression data sets and they demonstrate the advantage of our proposed shrinkage-based diagonal Hotelling's test. (C) 2015 Elsevier Inc. All rights reserved.

【 授权许可】

Free