UC San Diego

The regulation of expression states of genes in cells is a stochastic process. The relatively small numbers of protein molecules of a given type present in the cell and the nonlinear nature of chemical reactions result in behaviours, which are hard to anticipate without an appropriate mathematical development. In this dissertation, I develop theoretical approaches based on methods of statistical physics and many-body theory, in which protein and operator state dynamics are treated stochastically and on an equal footing. This development allows me to study the general principles of how noise arising on different levels of the regulatory system affects the complex collective characteristics of systems observed experimentally. I discuss simple models and approximations, which allow for, at least some, analytical progress in these problems. These have allowed us to understand how the operator state fluctuations may influence the steady state properties and lifetimes of attractors of simple gene systems. I show, that for fast binding and unbinding from the DNA, the operator state may be taken to be in equilibrium for highly cooperative binding, when predicting steady state properties as is traditionally done. Nevertheless, if proteins are produced in bursts, the DNA binding state fluctuations must be taken into account explicitly. Furthermore, even when the steady state probability distributions are weakly influenced by the operator state fluctuations, the escape rate in biologically relevant regimes strongly depends on transcription factor-DNA binding rates