- Computer simulation of
bio-molecular dynamics. We develop theory, algorithms, and computer code
to simulate the structure, dynamics, and thermodynamics of large
biomolecules. Our code is called MOIL.
- Long time dynamics.
Biophysical processes cover many orders of magnitude in time. These
processes are traditionally studied using many small time steps. For
example, 100000000000 steps are necessary to reach the time scale of
milliseconds, typical of many bio-molecular processes (e.g. channel gating);
a calculation that is not feasible today. We develop an approximate approach
to compute the long-time dynamics based on boundary value formulation that
makes it possible to use a very large step and still maintain the numerical
stability of the solution. We are working on extension of the methodology
that will enable calculation of time scales. We also develop the method of Milestoning
to compute the time evolution of the system based on reaction pathways,
approximate trajectories, or order parameters computed separately.
- Protein folding. We use
our novel long-time-dynamics algorithm to study protein folding. We
investigate the folding of protein
A, of cytochrome
c, and of a short alanine-rich helix
- Hemoglobin R to T transition. This is an on-going project to understand the
conformational transition in hemoglobin in which it changes its conformation
and its affinity to oxygen.
- Ligand diffusion in myoglobin. We are revisiting this
"classic" problem and are working on reproducing quantitatively
the rate of ligand diffusion from the protein.
- Myosin. We study the recovery stroke in this protein with