Development and application of computational methods to study nucleic acid structure and function

Molecular dynamics simulation is a powerful computational tool for studying biomolecular structure and function. Nucleic acid systems present special challenges since they are characterized by strong electrostatic forces and their structure and dynamics is sensitive to the surrounding ion environment. This dissertation is composed of a set of projects that, taken together, advance the state-of-the art modeling of nucleic acid systems. A common theme of this research is to provide a detailed atomistic description of RNA catalysis by applying a multi-scale modeling strategy. Computational RNA enzymology is complex in nature and requires consideration of many factors including direct and indirect interaction with ions, changes in residue protonation states, and conformational changes. This work aims to answer questions related to all three of these factors, which together with the chemical steps of the reaction, define the “problem space” of RNA enzymology. Chapters 3 and 4 represent a two step strategy for developing more balanced divalent ion parameters for nucleic acid simulations. In Chapter 3, we first evaluate the ability of 17 different Mg2+ non-bonded models to simultaneously reproduce structural, thermodynamic, kinetic and mass transport properties of the ion in solution. Chapter 4 introduces a set of newly developed Mg2+, Mn2+, Zn2+ and Cd2+ parameters that are accurate and balanced for interaction with nucleic acids, which are based on the Mg2+ model that performed best in aqueous solution in Chapter 3. In Chapter 5, molecular dynamics simulations are utilized to probe ion competition around nucleic acids and validate the ion models against experimental “ion counting” experiments. The effect of solvent model, salt conditions, and tether length on the structure and dynamics of a helix-junction-helix system is explored in Chapter 6. Chapter 7 concludes the thesis with a computational mutagenesis study of the glmS self-cleaving riboswitch conducted at several stages along the reaction path.Ph.D.Includes bibliographical referencesby Maria T. Pantev