Functional inter-species regularity can be sought by analyzing expression data collected from two organisms. The meaningful variation hinting at functional similarities and differences is hidden in the large and noisy data sets, and our task is to explore it.
A probabilistic associative clustering method is used for mining dependencies between the two data sets, gene expression data from orthologous mouse and human gene pairs. Associative clustering has been designed to be used in the first stage of analyzing a large data set. It presents the results as easily interpretable clusters which is a useful property for their biological interpretation.
The resulting clusters contain orthologous gene pairs with unexpected regularity between the expression profiles of the two species. Clusters with potential implications in biomedical research are chosen for a further study based on the biological interpretation of the results.