Exact solving and sensitivity analysis of stochastic continuous time Boolean models

摘要

One of the principle aims of systems biology is to understand with the help of models the complex molecular networks that regulate the functioning of a cell [ ]. To do so, numerous mathematical and computational formalisms have been used in the past decades [ ]. These range from quantitative and mechanistic models that require the knowledge of numerous biophysical constants [ ] to higher level, more qualitative models such as fuzzy logic [ ] and Boolean [ , ] models that describe functional dependencies, but not the details of biophysical mechanisms. Boolean models, originally introduced in the systems biology field by Kauffman [ – ], have the advantage that interactions between a model’s variables (that can be genes, proteins or other cellular components and their states) only need to be qualitatively defined and identifying attractors is a fast calculation [ ]. Traditionally, Boolean modeling has been used as a more qualitative approach to quickly identify the stationary states (attractors) of a model and test their robustness to initial conditions and/or perturbations. In most Boolean modeling platforms [ – ], time is described in discrete steps and model outputs are binary.