|
Activity Number:
|
428
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #304097 |
|
Title:
|
Multiple Comparison and Regularization for Analysis of eQTL Data
|
|
Author(s):
|
Jichun Xie*+ and Hongzhe Li
|
|
Companies:
|
University of Pennsylvania and University of Pennsylvania
|
|
Address:
|
5th Floor, Blockley Hall, Philadelphia, PA, 19104,
|
|
Keywords:
|
eQTL Data ; multiple comparison ; FDR control ; high-dimensionality
|
|
Abstract:
|
Experimental data have shown that the levels of gene expressions are inheritable. However, analysis of such data raises several interesting statistical problems. Due to high-dimensionality of both genotype and gene expression data and small sample sizes, controlling for multiple comparison has become important in order to achieve better power in detecting the genetic variants that are associated with the transcripts levels. We introduce an optimal procedure for FDR control in this setting. We will also introduce a regularization procedure for identifying the genetic variants that can explain the high-order expression modules. Simulations, analysis of real data sets and some theory will be presented.
|
