However, these genes also were not prognostic markers due to the low incidence of mutations in CRCs. Interestingly, additional factors that are neither metabolic sensors nor enzymes were shown to modulate BA homeostasis. As an alternative method of identifying reliable prognostic markers, we hypothesized that these factors may relay information on metabolic status between metabolic sensors and enzymes, functionally linking these two classes of molecules. These factors, called bridge proteins, may serve as reliable prognostic markers in patients with CRC, because anomalies in these proteins would disturb the delivery of metabolic information and the proper regulation of BA homeostasis. Current targeted approaches would be ineffective in probing relay proteins specifically between metabolic sensors and enzymes, due in large part to the lack of a method to quantify the relay degree of proteins. Systematic approaches, using information about known molecular interactions and the proteins connecting sensors and enzymes may identify and distinguish bridge proteins Epoxomicin 134381-21-8 implicated in cellular signaling networks. Here, we propose a network-based approach that identifies prognostic markers among proteins that play a critical role possibly linking sensors and enzymes of BA metabolism, relating to known biological hypothesis. These proteins, referred to as bridge proteins, can be assessed systematically based on information about molecular interactions recorded in several databases. To this end, we have defined a “bridgeness” metric, representing the degrees of connection between sensors and enzymes, and propose key bridge proteins as network markers for prognosis in patients with CRC. Using this “hypothesis-initiated” approach, we identified a set of markers that could better predict outcomes in patients with CRC than previously identified prognostic markers. A network-based investigation of biomarkers based on their bridgeness property may identify prognostic biomarkers implicated in cellular networks. By investigating genes involved in the regulation of BA homeostasis, this study has identified numerous genes for prognostic biomarkers of CRC, with showing mechanistic relevance to CRC pathogenesis. Although various prognostic biomarkers have been proposed based on biological hypotheses, these biomarkers have shown limited clinical usefulness. The hypothesis, that BAs play pivotal roles in CRC, provides clues to understanding the pathogenesis of this disease. However, rather than focusing on BAs themselves, we focused on the genes involved in regulating BA metabolism by linking metabolic sensors and metabolic enzymes. Based on a devised metric, “bridgeness”, numerous bridge proteins were selected from a reference, or bridge, network, and their prognostic abilities were analyzed. Bridge proteins could distinguish between normal and diseased tissues and are therefore relevant to the pathogenesis of CRC. These bridge proteins had greater and reproducible prognostic ability, as shown by statistical significance, than previously identified prognostic markers, suggesting that they are reliable prognostic markers in patients with CRC.