Validating gene clusterings by selecting informative gene ontology terms with mutual information

We propose a method for global validation of gene clusterings. The method selects a set of informative and non-redundant GO terms through an exploration of the Gene Ontology structure guided by mutual information. Our approach yields a global assessment of the clustering quality, and a higher level interpretation for the clusters, as it relates GO terms with specific clusters. We show that in two gene expression data sets our method offers an improvement over previous approaches.