|
Activity Number:
|
242
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #302027 |
|
Title:
|
A New Bi-Model Classifier for Predicting Outcomes of Prostate Cancer Patients
|
|
Author(s):
|
Zhenyu Jia*+ and Yipeng Wang and James Koziol and Michael McClelland and Dan Mercola+
|
|
Companies:
|
University of California, Irvine and The Sidney Kimmel Cancer Center and The Scripps Research Institute and The Sidney Kimmel Cancer Center and University of California, Irvine
|
|
Address:
|
36 Oxford, Irvine, CA, 92612, Department of Pathology and Laboratory Medicine, Irvine, CA, 92697,
|
|
Keywords:
|
Classifier ; Prostate Cancer ; Recurrence
|
|
Abstract:
|
The majority of prostate cancer cases are "indolent" that do not threaten lives. In order to improve disease management, reliable molecular indicators are needed to distinguish the indolent cancer from the cancer that will progress. Conventional methods can not apply to this study because the tissue samples are very heterogeneous in cell composition. Here we illustrate the expression level of any gene by a linear model considering the contributions from four principal types of cells and their interactions with aggression indicators (known relapse or nonrelapse status of the cases). ANOVA is used to identify cell specific relapse associated genes that possess discriminative power. The expression patterns of those selected genes may be described using two Gaussian models, one for relapse and one for nonrelapse cases. Thus they can be used for predicting outcomes of newly diagnosed cases.
|
