Gene set-based analysis of mucinous ovarian carcinoma

Abstract

ObjectiveMucinous ovarian carcinoma (MOC) is an uncommon subtype of epithelial ovarian cancers, and the pathogenesis is still poorly understood because of its rarity. We conducted a gene set-based analysis to investigate the pathogenesis of MOC by integrating microarray gene expression datasets based on the regularity of functions defined by gene ontology or canonical pathway databases. Materials and methodsForty-five pairs of MOC and normal ovarian tissue sample gene expression profiles were downloaded from the National Center for Biotechnology Information Gene Expression Omnibus database. The gene expression profiles were converted to the gene set regularity indexes by measuring the change of gene expression ordering in a gene set. Then the pathogenesis of MOC was investigated with the differences of function regularity with the gene set regularity indexes between the MOC and normal control samples. ResultsThe informativeness of the gene set regularity indexes was sufficient for machine learning to accurately recognize and classify the functional regulation patterns with an accuracy of 99.44%. The statistical analysis revealed that the GTPase regulators and receptor tyrosine kinase erbB-2 (ERBB2) were the most important aberrations; the exploratory factor analysis revealed phosphoinositide 3-kinase-activating kinase, G-protein coupled receptor pathway, oxidoreductase activity, immune response, peptidase activity, regulation of translation, and transport and channel activity were also involved in the pathogenesis of MOC. ConclusionInvestigating the pathogenesis of MOC with the functionome provided a comprehensive view of the deregulated functions of this disease. In addition to GTPase regulators and ERBB2, a plenty of deregulated functions such as phosphoinositide 3-kinase, G-protein coupled receptor pathway, and immune response also participated in the interaction network of MOC pathogenesis.