Seminario Prof. Ernst Wit - 21 maggio 2003

[Patrizia Piacentini <piacent@stat.unipd.it>]

     UNIVERSITA' DEGLI STUDI DI PADOVA
    DIPARTIMENTO DI SCIENZE STATISTICHE
      Via C. Battisti, 241 - Padova

******************************************
 
               Seminario

  THE USE OF BAYESIAN HIERARCHICAL MODELS 
          IN MICROARRAY STUDIES
 
             Prof. ERNST WIT
        Department of Statistics
          University of Glasgow


             21 maggio 2003
               ore 15:30
              Aula CUCCONI

******************************************

Abstract

Currently the study of data from microarray experiments is subject
to a lot of very necessary but low-level analysis on the level of
data quality and normalization. Actual inference from these data
seems often limited to hypothesis tests for differential expression
or at hoc clustering of similarly expressed genes. In this talk I
want to set out a single methodology, namely Bayesian hierarchical
modelling, for approaching inferential problems and suggest that we
can push inference forward by considering additional covariate
information. I present the example of a dynamical model for time
series expression data observed on the circular tuberculosis genome.



**********************************************
Patrizia Piacentini
Segreteria  Organizzativa  Convegni e Seminari
Dipartimento di Scienze Statistiche
Università degli Studi di Padova
Via C. Battisti, 241
35121 Padova

tel 049 8274167
fax 049 8274170
e-mail segrorg@stat.unipdit
***********************************************