Improved Methods of Simulation and Analysis For Stochastic Processes in Cell Biology

Stochasticity (that is, randomness) is an inherent property of many biological systems. For example, gene expression is stochastic, resulting in random fluctuations of mRNA and protein copy numbers in the cell. In cell differentiation, there is evidence that the phenotype of the cell can be driven toward an entirely different type of cell due to noise. Stochastic fluctuations are also important in the spatio-temporal dynamics of molecular interactions within the cell, affecting processes such as cell activation and signal transduction. To gain a better understanding of biological systems, computer simulations of biomolecular processes in the cell are increasingly utilized to complement experiments, quantify mechanistic hypotheses, and predict the effect of perturbations. Stochastic models, in particular, can be prohibitively expensive to simulate and difficult to analyze. In this work, we develop and extend methods of stochastic simulation and analysis that are applicable to a variety of cell biological systems. We focus on two specific application areas: The first is development of a method to analyze gene regulatory network models that have multiple, metastable states. The method enables a simplified, quantitative representation of complex phenotype landscapes and transitions. Second is the development of improved simulation methods for spatial stochastic systems. This work focuses on rare events in reaction-diffusion systems and found several extensions to currently-employed simulation methods which improve simulation efficiency