Empirical Bayes analysis of unreplicated microarray data (Q964653)

A microarray analysis technique is developed for the analysis of unreplicated gene expression data. It is based on a two-way ANOVA model with heterogeneous errors. A Bayesisan approach is used for the parameter estimation with Markov chain Monte-Carlo techniques for the evaluation of the estimates. An empirical Bayes approach is proposed for the data variance priors specifications. The performance of the proposed methods is assessed by real data and via simulations.