Title

Bayesian Models For The Multi-Sample Time-Course Microarray Experiments

Keywords

Bayesian analysis; Classification; Hypothesis testing; Multisample problems; Time-course microarray

Abstract

In this paper we present a functional Bayesian method for detecting genes which are temporally differentially expressed between several conditions. We identify the nature of differential expression (e.g., gene is differentially expressed between the first and the second sample but is not differentially expressed between the second and the third) and subsequently we estimate gene expression temporal profiles. The proposed procedure deals successfully with various technical difficulties which arise in microarray time-course experiments such as a small number of observations, non-uniform sampling intervals and presence of missing data or repeated measurements. The procedure allows to account for various types of errors, thus, offering a good compromise between nonparametric and normality assumption based techniques. In addition, all evaluations are carried out using analytic expressions, hence, the entire procedure requires very small computational effort. The performance of the procedure is studied using simulated data. © Springer-Verlag 2012.

Publication Date

12-31-2012

Publication Title

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Volume

7548 LNBI

Number of Pages

21-35

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

DOI Link

https://doi.org/10.1007/978-3-642-35686-5_3

Socpus ID

84871596545 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/84871596545

This document is currently not available here.

Share

COinS