Enabling Tools for the Mechanical Analysis of Protein Molecules
M.S. Defense Announcement For
Date: Friday, July 27, 2007
Time: 9:00 a.m.
Location: ITEB 360
Major Advisor: Prof. Ilies
Co-Major Advisor: Prof. Kazerounian
Associate Advisor: Prof. Fan
The ability to predict the changes in the protein conformations when subjected to external stimuli is one of the open and most significant problems in science today. This thesis investigates three important topics related to this problem, and relies on the assumption that proteins can be modeled as kinematic chains. Specifically, we present the development and implementation of protein visualization techniques, and of two approaches to perform side chain level inverse kinematics, as well as a comparison of two energy reduction methods for protein structures.
First, several visualization methods were developed and implemented for both protein structure representation, as well as data visualization, and have been used in all subsequent work in this thesis. Second, the efficiency, robustness and convergence of multibody dynamic simulation, and of our own kinetostatic compliance framework have been evaluated. We show that both methods are effective at reducing the potential energy of a protein structure, and that the kinetostatic compliance method is not only the more efficient, but also the more robust of the two methods. Finally, we present and compare two methods for calculating joint angles that match a rigid body protein model to atomic coordinates provided by the Protein Data Bank. These angles, are essential for improving the protein folding simulation within the kinetostatic compliance framework.
Biographical Sketch of Candidate: Jesse Parker received is B.S. in Mechanical Engineering at Montana State University in 2004, and began work with Dr. Ilies and Dr. Kazerounian spring of 2005. He has been partially supported as an NSF fellow for the Galileo program, working inlocal schools as an engineering educator.