|
Activity Number:
|
467
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
ENAR
|
| Abstract - #305050 |
|
Title:
|
Testing for Gene Effect in Presence of Gene-Gene Interactions
|
|
Author(s):
|
Arnab Maity*+ and Xihong Lin
|
|
Companies:
|
Harvard School of Public Health and Harvard School of Public Health
|
|
Address:
|
Department of Biostatistics, Boston, MA, 02115,
|
|
Keywords:
|
High dimensional data ; Semiparametric Regression ; Kernel machine regression ; Score test ; Garrote kernel
|
|
Abstract:
|
Our interest lies in testing for gene effect in the presence of possible gene-gene interactions. In general, testing for the effect of a particular gene in the presence of gene-gene interaction requires testing for the corresponding interaction effect of the gene of interest with other genes. However, testing for all possible interactions requires a strong parametric assumption and is likely to lead to loss of power. We develop score based tests for main effects using kernel machine technique that do not require testing for the coefficients of all the interaction terms and thus allowing us to obtain more powerful tests using less degrees of freedom. We will investigate the asymptotic properties of our test, evaluate its performance via simulation studies and apply it to the Michigan prostate cancer data.
|
