Coupling regulatory networks and microarays: Revealing molecular regulations of breast cancer treatment responses

Abstract

Moving towards the realization of genomic data in clinical practice, and following an individualized healthcare approach, the function and regulation of genes has to be deciphered and manifested. Two of the most significant forms of molecular data come form microarray gene expression sources, and gene interactions sources - as encoded in Gene Regulatory Networks (GRNs). The usual computational task is the gene selection procedure with the GRNs to be mainly utilized for annotation and enrichment purposes. In this study we present a novel perception of these resources. Initially we locate all functional path-modules encoded in GRNs and we try to assess which of them are compatible and match the gene-expression profiles of samples that belong to different phenotypes. The differential power of the selected path-modules is computed and their biological relevance is assessed. The whole approach was applied on a set of microarray studies with the target of revealing putative regulatory mechanisms that govern and putatively guide the treatment responses of BRCA patients. The results were quite satisfactory according to their biological and clinical relevance. © 2012 Springer-Verlag.