Comprehensive analysis of cancer-stromal interactome

Cancer-microenvironment interaction is an important target for cancer therapy. However, finding new druggable interactions is challenging due to the lack of quantitative methods to analyze whole cancer-stromal interactome. We propose CASTIN (CAncer-STromal INteractome analysis), a novel system for the evaluation of cancer-stromal interactome from RNA-seq data using cancer xenograft models. Three interactome evaluation indices coupled with curated ligand-receptor interaction database in our system provide quantitative and comprehensive view of interactome and enable the identification of critical cancer-microenvironment interactions. We applied the system to pancreas cancer dataset and successfully characterized the individual tumor in terms of cancer-stromal relationships, and identified well-known druggable interactions.