A Perturbation Approach for Refining Boolean Models of Cell Cycle Regulation

Considerable effort is required to build mathematical models of large protein regulatory networks. Utilizing computational algorithms that guide model development can significantly streamline the process and enhance the reliability of the resulting models. In this article we present a perturbation approach for developing data-centric Boolean models of cell cycle regulation. We assign a score to a network based on the steady states of the network-dynamics, and the dynamical trajectories corresponding to the initial conditions. Then, perturbation analysis is used to find new networks with lower scores, in which dynamical trajectories traverse through the correct cell cycle path with high frequency. We apply this method to refine Boolean models of cell cycle regulation in budding yeast and mammalian cells.