NETWORK INFORMATION FLOW REVEALS KEY GENES IN GLIOBLASTOMA PROGRESSION

Regulatory networks within the cell are organized into a multilayer architecture with feed-forward and feedback control motifs. Current efforts to understand how information is propagated in the cell are confined to a monolayered view or omit feedback loops that are key features of many regulatory networks. Here we present an information flow algorithm that simulates propagation of regulatory signals over a multilayer representation of cellular networks. Through integration of multiple data types and interaction rules inspired by Google’s webpage ranking algorithm, we show using the TCGA glioblastoma multiforme (GBM) mRNA / miRNA dataset as well as previously compiled regulatory relationships among TFs, miRNAs and target genes that this novel algorithm is able to uncover key regulators during cancer progression.