Activating Cancer Hallmarks through Changes in mRNA/Protein Regulation

As a diverse family of diseases, cancer is unified by a set of common dysfunctions, such as limitless growth potential and an insensitivity to antigrowth signals. These shared overarching biological processes have been termed the hallmarks of cancer. To better understand the root cause of cellular dysregulation, intense molecular characterization of tumors has utilized DNA, RNA, and protein measurement techniques to produce proteogenomic data. In large cancer cohort studies, genomic and proteogenomic data have frequently identified many cancer hallmarks including cell cycle and cell signaling. However, altered metabolism, a known cancer hallmark, is not as clearly identified in mutation screens or differential expression analyses. Here, we introduce a new computational method to identify changes in cellular regulation by focusing on the mRNA/protein relationship. We create a metric, Δ_corr, to capture when the mRNA/protein correlation changes significantly between tumor and normal tissues and show that it is distinct from differential expression and also not associated with DNA mutation profiles. Our method clearly highlights altered metabolic pathways across multiple tumor types. Δ_corr gives researchers a new perspective on the dysfunction of tumor cells and introduces a novel method for proteogenomic data integration.