Cells co-regulate the expression of genes forming modules to respond to environmental stimuli. Depending on the stimulus, these modules contain different genes, have distinct temporal response patterns or are left unaffected. Today, an increasing pool of microarray datasets monitoring gene expression over time exists. These gene-time datasets can be searched for coregulated gene expression modules by one of the many clustering methods published so far.

However, for an integrative analysis, several gene-time datasets can be joined into one three-dimensional gene-condition-time dataset, to which standard clustering or biclustering methods, however, cannot be applied. In this work we present a novel probabilistic clustering approach for gene-condition-time datasets.