Mathematics - Colloquium
Friday, February 26, 2010
11:00 AM-12:00 PM
ABSTRACT: Developing probabilistic models for modeling regulatory networks based on genome-wide high-throughput data is becoming more and more popular in biostatistics and bioinformatics. Bayesian network and Gaussian graphical models are the two most widely used methods. Our research is mainly focused on using Bayesian network models to discover gene networks based on their large-scale microarray expression data. We use mutual information to measure the dependency among genes. A three phase procedure is proposed with the outcome being a gene relation map. Our method is a modified version of Cheng et al (2002). Our proposed procedure is implemented using R. We also briefly discuss an idea of network based classification using our gene relation map.
Suggested Audiences: Adult, College
E-mail:
ma-chair@wpi.edu
Last Modified: February 17, 2010 at 4:08 PM
Powered by the Social Web - Bringing people together through Events, Places, & Common Interests