Identification of heterogeneous cancer mechanisms by using modular gene expression patterns

Abstract

Cancer is known for its heterogeneous clinical behaviors and various patient outcomes. For the understanding of the complexities of the disease mechanisms and to develop more efficient therapeutic strategies, the ability to dissect this heterogeneity and to identify subgroups representing the common cancer mechanisms is crucial. In this study, we used an integrative transcriptome analysis to identify transcriptional modules as functional base of breast cancer mechanisms. We identified gene modules that shared co-expressed patterns in a subset of patients across multiple datasets. The composite functions of gene modules were inferred. We show that the identified subgroups share patterns of module activity and exhibit clinical and biological properties. Our results showed molecular bases of breast cancer for improved predictor and classifier in clinical use.