Department of Statistics Seminar
North Carolina State University
presents
University of British Columbia
"Gene classification in deletion set studies"
ABSTRACT
Deletion set studies are a powerful tool for gaining insight into the functional roles of individual genes. The yeast deletion set (YDS) is a large collection of yeast strains, each characterized by the absence of exactly one gene. By studying the response of the YDS to various treatments, we can obtain functional information about individual genes. We propose a statistical method to extract functional information from deletion set studies and to provide appropriate measures of confidence. This approach is based on regression analysis and the semiparametric bootstrap and is a generalization of methods developed by the speaker and Mark van der Laan (UC Berkeley) for estimating gene subsets and clusters using gene expression data from microarrays. The biological motivation and context is an ongoing collaboration between the speaker (UBC, Statistics and Biotechnology Lab), Kristin Baetz (UBC, Centre for Molecular Medicine and Therapeutics, Phil Hieter Lab), and Michel Roberge (UBC, Biochemistry). The broad research goal is to identify specific mechanisms by which Motuporamine-C, a compound isolated from a sea sponge, inhibits angiogenesis in human cancer cells in vitro. The immediate focus is to obtain preliminary information in the model organism S. cerevisae
Friday, November, 08, 2002
3:35 - 4:35 pm
206 Cox Hall
Refreshments will be served on the second floor of Dabney Hall (left of Room 222) at 3:00 pm.