Parameterization of molecular mechanics (MM4) for large biological systems: Small peptides - emphasis on protein modeling, Alkylphosphines - emphasis on DNA structure.

摘要

The use of molecular mechanics programs to study the structure, vibrational frequencies, dipole moments, and moments of inertia of biological compounds is becoming more common. Due to the limitations of ab initio calculations, properly parameterized molecular mechanics calculations are the only accurate method available for the study of large compounds. This dissertation details the results from three separate studies of the use of the Allinger molecular mechanics force field for biological systems.;Chapter 2 involves a comparison of dipole moments of forty-four organic molecules. The results show MM3 (2000) to be the best method at predicting the magnitude of the dipole moment. The direction of the dipole moment was also compared and all of the computational methods give similar directional predictions for the dipole moment of each molecule.;Chapter 3 deals with the parameterization of phosphine and the alkylphosphines for Molecular Mechanics, MM4. This parameterization work is the first step in allowing MM4 to properly model nucleotides and phosphine moietys.;Chapter 4 involves the parameterization of MM4 for amides and peptides. By mapping the DFT torsion curves, MM4 is able to accurately model small peptides. A set of twelve amino acids were also studied and the results show that MM4 is better then MM3 at predicting molecular structures when compared against the DFT calculated result.;The results of this dissertation support the use of molecular mechanics as a valid and important computational technique.