Protein folding dynamics in lattice model with physical movement.

We study folding dynamics of protein-like sequences on square lattice by constructing a physically realizable move set that exhausts all possible conformational changes for a structure. By solving the master equation of 16-mer characterized by a 802,075times802,075 transition matrix, we monitor the time-dependent probabilities of occupancy of all conformations over 9-orders of time scale from the first kinetic move until reaching Boltzmann equilibrium. We find that folding rates of protein-like sequences adopting the same ground state conformation differ as much as 200 times, and parameters of the native structures, designability, and thermodynamic properties are weak predictors of the folding rates in our model systems. Instead, we show that properties of the kinetic energy landscape defined by the connection graph of physical moves can provide excellent account for observed folding rates. Without the approximation of macrostates, we show how transiently accumulating intermediate states can be identified by basin analysis of the kinetic energy landscape.