SEMINARS
|
|
Spring 2007
STATISTICS
COLLOQUIUM
Wednesday, March 28, 2007
3:30-4:00—Refreshments
4:00-5:00—Talk
Yost Hall, Room 101
Yuping Wu, PhD
Department of Mathematics, Cleveland State University
Feature-Based Classification of Proteomics Data
Abstract
The use of mass spectrometry (MS) as a means of analyzing the proteome
for biomarker discovery and cancer diagnostics has been evaluated
extensively in recent years due to the new advances in MS technology.
The primary goals of this project are development of statistical
methods for discriminant analysis of proteomics data and disease
biomarker discovery. In this talk, I will emphasize peak extraction
from a baseline properly corrected spectrum, prediction of disease
status for a new sample, and discovery of biomarkers based upon
location/intensity information associated with peaks. The methods
will be evaluated via applications to various forms of existing
proteomics data, available in the literature and in the public
domain. A series of computer-intensive Monte Carlo Simulations will
assess objectively the operating characteristics of the new
methodology.
|
