The Liang laboratory uses molecular simulations to fundamentally understand how electronically non-
adiabatic reactions couple with protein’s structure, dynamics, and function. Electronically non-adiabatic reactions,
such as photochemical and electron transfer reactions, switch electronic states during the chemical
transformation. A fundamental understanding of how they interact with proteins is essential for advances in
biomedical sciences. However, two central and fundamental questions remain elusive: (1) how does the protein
environment modulate the pathway, dynamics, and quantum yields of the non-adiabatic reactions? (2) how do
the non-adiabatic reactions induce structural changes in the protein? Molecular simulation is indispensable to
answering these questions because it can resolve the energetics and kinetics of chemical reactions at atomic-
level detail, which is often beyond the limit of current experimental techniques. Also, simulation incurs minimal
cost and has no risk for human subjects. However, the multiscale nature of these processes poses significant
challenges for traditional computational methods. Specifically, standard molecular mechanics (MM) simulations
cannot describe the quantum-mechanical (QM) nature of the non-adiabatic reactions. Meanwhile, typical QM
simulations are too expensive to characterize the slow biomolecular motions in response to these reactions.
To overcome these challenges, in the next five years, our research program will expand our current efforts
to develop and employ multiscale simulation methods to understand (1) the light-regulated signaling activities of
transient receptor potential channels and metabotropic glutamate receptors by synthetic molecular switches,
which are of top interest in optogenetics and photopharmacology, and (2) the long-range electron transfer events
in cryptochromes and electron bifurcating enzymes, which are fundamental to understanding the circadian clocks,
magnetic field sensing and energy metabolism in living organisms. The unique advantages of our approaches
include (1) accurate and efficient non-adiabatic dynamics simulations with “on-the-fly” ab initio calculations of
nuclear gradients and electronic couplings; (2) effective integration of the high-quality non-adiabatic dynamics
simulations with high-efficiency MM sampling of protein conformational change. These key methodological
advantages will enable the comprehensive characterization of non-adiabatic chemical reactivity in complex
biomolecular systems and answer the above-mentioned fundamental questions with unprecedented accuracy.
Explicitly simulating the photodynamics of biomolecules of this size and complexity is not routine, especially
with the proposed multiscale simulation framework that incorporates ab initio non-adiabatic dynamics simulations.
Therefore, five years into the future, our research will provide new insights into the design principles of next-
generation photochemotherapy with minimal side effects, create powerful computational tools for simulating
electron transfer in biomolecules, and deepen our fundamental understanding of the roles of quantum mechanics
in biology in general.