|
Activity Number:
|
546
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #306596 |
|
Title:
|
Effects of Dependencies in High-Dimensional Multiple Testing Problems
|
|
Author(s):
|
Kyung In Kim*+ and Mark A. van de Wiel
|
|
Companies:
|
Eindhoven University of Technology and Eindhoven University of Technology
|
|
Address:
|
P.O. Box 513, Eindhoven, 5600 MB, The Netherlands
|
|
Keywords:
|
gene expression data ; multiple testing ; conditional independence
|
|
Abstract:
|
We consider effects of dependencies among variables of high-dimensional data in multiple testing problems. Recent simulation studies considered only simple correlation structure among variables, which was hardly inspired by real data features. Our aim is to describe dependencies as a network and systematically study effects of several network features like sparsity and correlation strength. We discuss a new method for efficient guided simulation of dependent data, which satisfy the imposed network constraints. We use random correlation matrices and perform extensive simulations under nested conditional independence structures. We check the robustness against dependency of several resampling-based methods. Powers computed from popular methodologies such as Benjamini-Hochberg FDR, SAM and mixture models are compared. Finally some applications to gene expression data are illustrated.
|
