Orthogonal space sampling of protein functional dynamics

Proteins aimlessly fluctuate and occasionally are activated through specific energy flow channels for functional transitions. Such functional conformational transitions occur at timescales much longer than affordable by current computing power. A long-standing question has been how to accelerate all atom molecular dynamics simulation faithfully along physical energy flow pathways. For this paper, orthogonal space sampling theory was formulated, beginning with the orthogonal space random walk and orthogonal space tempering methods. In this talk, the complete Hamiltonian of high-order orthogonal space tempering (HOOST) will be presented with successful application of predictive sampling of long-timescale protein functional dynamics.