Protein modeling is becoming increasingly important for the characterization of biological systems at the molecular level. It is now widely accepted that the key to understanding biological function, and subsequently disease, is to understand the structure, flexibility, kinetics, and dynamics of the body's worker molecules, proteins. Recent successes in understanding the structure and dynamics of proteins using computational analysis underscore the importance of developing computational methodologies to explore the fundamental principles of protein flexibility. Representing a protein by a single configuration is less and less attractive, as such a representation does not represent the dynamical interplay of different conformations that can regulate protein function and does not fully characterize the protein's interaction with neighboring molecules.
This proposal will develop a computational framework for the characterization of protein flexibility at equilibrium conditions through a combination of geometric/robotic and biophysics methods. The proposed work lies at the intersection of computer science and modern biophysics, and will benefit both communities. On the computational side, it will lead to new methodologies and paradigms to model physical systems with high flexibility, complex geometry, multiple constraints, and continuous motion. Successful robotics and computational geometry methods will be adapted to support the large-scale analysis required for biological applications. On the biophysical side, novel theories and methodologies will be developed and tested for modeling proteins at different resolutions. Quantitative connections between theory and experiment will be pursued. The proposed work has the potential to dramatically affect major unresolved problems on the inner workings of protein systems.
