Computer Simulations of Dynamical Processes in Biomolecular Systems

This dissertation presents the application of computer-simulation techniques to several problems in biochemistry and biophysics. First, we consider a single-species diffusion-limited annihilation reaction confined to a two-dimensional surface with one arbitrarily large dimension and the other comparable to in size to interparticle distances. This situation could describe reactants which undergo longitudinal and transverse diffusion on long filamentous molecules (such as microtubules), or molecules that undergo one-dimensional diffusion (e.g. a transcription factor on DNA) but simultaneously exhibit diffusive behavior in a second dimension corresponding to a rotational or conformational degree of freedom. We combine analytical arguments and Monte Carlo simulations to show that the reaction rate law exhibits a crossover from one-dimensional to two-dimensional diffusion as a function of particle concentration and the size of the smaller dimension.;We also use Monte Carlo simulations to study the assembly of viral capsids around nucleic acids or other polymers. We simulate on a lattice the dynamical assembly of closed, hollow shells composed of hundreds of subunits around a flexible polymer. Assembly is most efficient at an optimum polymer length that scales with the surface area of the capsid; longer-than-optimal polymers often lead to partial-capsids with unpackaged polymer "tails" or a competition between multiple partial-capsids attached to a single polymer. The polymer can increase the net rate of subunit accretion to a growing capsid both by stabilizing the addition of new subunits and by enhancing the incoming flux of subunits. Some of these predictions have been confirmed by experiments.;Molecular dynamics simulations of proteins at atomic resolution are a powerful tool for discovering pathways of conformational change. We consider a simulated pathway between the inactive and active states of a signaling protein, nitrogen regulatory protein C, and demonstrate that the loss of native stabilizing contacts during activation is compensated by non-native transient atomic interactions during the transition. These findings are tested and confirmed by experimental measurements of mutated proteins. We also generate realistic simulated pathways for the conformation change in the enzyme adenylate kinase and propose molecular mechanisms which distinguish pressure-sensitive and pressure-insensitive variants of the protein.