The first use of molecular dynamics in studying a biological process was presented by Warshel in 1976 work . His surface hopping simulations of the primary event in the vision process predicted correctly the photoisomerization time of rhodopsin. This prediction has been confirmed experimentally and provides an illustration of the predictive power of microscopic simulations. The same semiclassical approach has been used in studies of the primary electron transfer in photosynthesis.   
Our group continues to work on the excited state dynamics of bactreiorhodopsin and of GFP, as well as fundamental studies of light induced surface crossing conical intersections. See also movie.