Simulation Algorithms and Multiscale Modeling

Many biomolecular functions can’t be directly observed, even with the state-of-the-art experiments. Instead, we can simulate these biomolecules on a computer using what we know about physics. But, even simple biomolecular systems can have billions of atoms and we might want to study them over a period of milliseconds or seconds—a big challenge when we need to solve Newton’s equations of motion every femtosecond. Luckily, there are some clever things we can do with statistical mechanics to make this computationally tractable.

I am interested in designing new multiscale algorithms for this purpose. My goal is reach new problems in biology; those that can't be tackled with traditional, periodic simulations. Two particular advances I'm excited about are: (1) adaptive boundaries that shrink and expand with respect to an evolving solute or a growing crystal/molecular assembly (see Wagoner and Pande JCP 2018 or this example for peptide folding); (2) Grand canonical control of not just solvent, but ligands and other ions (see an example for micelle formation). These algorithms are both theoretically rigorous (they sample the correct stationary distribution) and accurate.

Nonequilibrium Stastistical Mechanics

Life is a nonequilibrium phenomenon and, at its core, all of biology is driven by the energy dissipation and entropy production of living systems. I use nonequilibrium theory to understand how biological systems like the cell are driven by energy dissipation and entropy production. Protein synthesis, ion and pH homeostasis, kinetic proofreading, and cell fate determination are all nonequilibrium phenomena. If the principles of Cell Biology tell us what happens, Nonequilibrium Physics tells us why and how.

The Evolution and Design of Biomolecular Machines

In one day, an average person will synthesize, use, and recycle his or her body weight in ATP (a 2000 calorie diet roughly equals 50 kg of ATP). All of this ATP must be synthesized by F0F1-ATPase, a biomolecular motor. Evolution places extraordinary pressures on F0F1-ATPase to be a fast and efficient machine.

For these reasons, molecular motors like F0F1-ATPase are ideal for studying the overlap of thermodynamic driving forces and biological evolution. I am particularly interested in the nonequilibrium driving forces that are present both in the mechanisms of individual motors (like F0F1-ATPase) and in the communication among sets of motors (like the myosin II motors that coordinate their actions during muscle contraction).