Bayesian network classifiers for time-series microarray data

Abstract

Microarray data from time-series experiments, where gene expression profiles are measured over the course of the experiment, require specialised algorithms. In this paper we introduce new architectures of Bayesian classifiers that highlight how both relative and absolute temporal relationships can be captured in order to understand how biological mechanisms differ. We show that these classifiers improve the classification of microarray data and at the same time ensure that the models can easily be analysed by biologists by incorporating time transparently. In this paper we focus on data that has been generated to explore different types of muscular dystrophy. © Springer-Verlag Berlin Heidelberg 2005.