Activity Number:
|
265
|
Type:
|
Invited
|
Date/Time:
|
Wednesday, August 14, 2002 : 8:30 AM to 10:20 AM
|
Sponsor:
|
International Chinese Statisticial Association
|
Abstract - #300238 |
Title:
|
Modeling the Influence of Disease Processes on Gene Expression
|
Author(s):
|
Kirby Shedden*+
|
Affiliation(s):
|
University of Michigan
|
Address:
|
, Ann Arbor, MI, ,
|
Keywords:
|
|
Abstract:
|
I will discuss two statistical models that aim to characterize the influence of tumor progression on gene expression. The goals are (i) to uncover variant trajectories of tumor progression, (ii) to identify marker genes that are associated with different stages of tumor development, and (iii) to predict the development status of individual tumors. Both models are based on viewing tumor development as causing monotonic alteration of gene expression relative to expression in normal cells. The first model is a simple approach using least squares fitting and dimension reduction. The second model is a hierarchical Bayesian model that incorporates ideas from mixture modeling and Bayesian variable selection. The notion of variant progression trajectories provides a clustering of the samples that can be viewed as a biologically-motivated analogue to data-driven clustering procedures such as hierarchical clustering.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2002 program |