Spring Seminar Series - January 23, 2003
University of Minnesota
School of Statistics
College of Liberal Arts
Dimension reduction methods in the analysis
of genomic data
Francesca Chiaromonte
Statistics Department
Pennsylvania State University
Thursday, January 23, 2003
4:00 PM, 115
Ford Hall
Minneapolis, East Bank Campus
Social at 3:30 PM, 300
Ford Hall
Abstract
Traditional dimension reduction methods such as
principal components, and more modern approaches such as sufficient dimension
reduction, can play a crucial role in the analysis of large data sets produced
by DNA sequencing, DNA alignment, and global gene expression recording techniques
(e.g. microarrays).
In this talk, I will briefly review inverse regression methodology for
sufficient dimension reduction (Sliced Inverse Regression, Partial Sliced
Inverse Regression) and present some applications that exemplify broad areas
of recent bioinformatics research.
These include the discrimination of genomic regions performing biological
functions based on DNA alignment information, the characterization of genes
in terms of DNA sequence information, and the discrimination of cancer types
based on global gene expression information.