| Using traditional clustering in microarray analysis, one can now identify homogeneous groups of genes that co-express similarly. Parallel gene monitoring becomes essential to understand gene expression regulatory mechanisms, especially in the case of diseases. Furthermore, recent studies introduced subspace clustering using the Gene Ontology, yielding clusters of genes with similar biological function. In our work we propose to combine these two methods to investigate genes involved in apparition of breast cancer. We believe that isolating subclusters of genes similar both in expression patterns and functions may provide valuable information on their regulation. |