Title: "Bayesian Approach for Identifying Molecular Signatures"
Speaker: Min Zhang, Department of Statistics, Purdue University
Place: Mechanical Engineering (ME) 161; October 2, 2007, Tuesday, 4:30pm

Abstract

Recently, much effort has been devoted to identifying cancer specific molecular signatures for early diagnosis, monitoring effects of treatment, and predicting patient survival time. The statistical analysis is challenged by the large number of predictors but a small number of observations. We proposed a two-stage procedure to profile molecular signatures for survival outcomes. The performance of our approach is demonstrated with a simulation study. We applied it to a diffuse large B-cell lymphoma study and identified some new candidate signatures for disease prognosis.

Associated Reading: M. Zhang and D. Zhang. 2007. Bayesian profiling of molecular signatures to predict event times. Theor Biol Med Model. 2007; 4:3.



Click here for a full schedule of BIOINFORMATICS SEMINARS, past and present.