New computational tools for macromolecular simulation and modeling: Charge assignment, structure modeling, and molecular dynamic simulation

In this dissertation, we described the work that we have done in three major subjects that deal with the development of macromolecular simulation and modeling. First, we described a method for predicting charge distributions in molecules, which can be used for both calculation of solvation free energy by finite-difference Poisson Boltzmann method as well as charge prediction in molecular dynamic simulations. The assigned partial charges can reflect the effects of the conformational changes and solvent induction on charge distribution in molecules. Second, we reported a new methodology for rapid generation of alternative protein structures based on a modified elastic network model. The generated structures (1) Differ significantly from the starting structure. (2) Have good stereochemistry (bonded geometry) and good steric properties (absence of atomic overlap). We show that this methodology is also capable of modeling transition pathways and providing alternative templates and models in comparative modeling. Third, we studied the molecular dynamics of RNA tetraloop molecules in a newly developed explicit ion, implicit water solvent simulation model and compared the simulation performance with the explicit MD simulation. The implicit water model uses the finite difference Poisson model with the smooth permittivity method implemented in the ZAP libraries. The entire ion/water treatment is interfaced with the molecular dynamics package CHARMM. Using the methodology, the implicit solvent model produced stable simulations on RNA molecules with structures remaining close to experiment.