|
Activity Number:
|
63
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #310370 |
|
Title:
|
Penalties Paid for Ignoring the Correlation Between Intensity Channels in Gene-Expression Analysis
|
|
Author(s):
|
Sang-Hoon Cho*+ and Michael Kosorok and Richard A. Johnson
|
|
Companies:
|
University of Wisconsin-Madison and The University of North Carolina at Chapel Hill and University of Wisconsin-Madison
|
|
Address:
|
MSC 1300 University Ave, Madison, WI, 53706,
|
|
Keywords:
|
microarray ; bivariate gamma distribution ; copula function ; Bayes estimator
|
|
Abstract:
|
In a typical two channel microarray analysis, it is natural to expect that there exists a strong positive correlation between intensity channels, especially under the null hypothesis of no differentially expressed genes. As far as we are aware, there has not been any study specific investigating about how much gain there would be once the correlation information between intensity channels was utilized in statistical approaches. In this talk, we will show that the correlation information between intensity channels may be a significant component embedded in the data that should not be overlooked in gene-expression analysis. As an illustration, we will show a couple of examples in which we specify a correlation component in statistical methods proposed by Chen et al. (1997) and Newton et al. (2001) and make a comparison between the earlier models and our extended models with correlation.
|
