A mathematical model of biological signaling networks and network characteristics correlated with signaling behavior.

摘要

Traditionally, molecular biology has focused on the role of individual genes. More recently, systems biology has shifted the focus to interactions among many genes; the field emphasizes that the behavior of genetic networks is important and difficult to predict from the knowledge of a single gene. This work studies interacting biochemical networks. In particular, we focus on the characteristics of signaling networks. Biological signaling occurs when a chemical outside the cell (the signal) binds to a receptor on the surface of the cell. This causes a signaling cascade of chemical reactions in the cell, leading to a change in cellular behavior. When a cell does not properly respond to its signals, cancer or other diseases can result.; We developed a simplified dynamical systems model to describe cellular signaling. The model is based on a model of interacting genetic networks (developed by Wagner and extended by Siegal and Bergman). One element of the system's state vector is identified as the signal. The influence of the signal on other elements of the network allows the system to switch between different stable steady states depending on the state of the signal. Using our model and mathematical definition of signaling, we studied the network characteristics associated with signaling behavior in small networks (2, 3, or 4 elements). We find that the most important parameters associated with signaling behavior are the structure of the network and the number and placement of non-zero connections between elements. The more connections there are from the signal to the subnetwork (the network with the signal and connections to/from the signal removed), the more likely the network is to signal. Networks that signal are not likely to be full rank. In addition, self connections, particularly negative self connections, are suppressed in signaling networks, compared to the full population of networks. Finally, we use our model to study an example biological signaling system (a phosphotransfer signaling pathway). This work gives insight into the network structure that would most readily allow cells to evolve signaling behavior.