Probing the energy landscape of alanine dipeptide and decalanine using temperature as a tunable parameter in molecular dynamics

We perform several molecular dynamics (MD) calculations of solvated alanine dipeptide and decalanine in vacuum with temperature as a tunable parameter and in the process, generate Markov state models (MSMs) at each temperature. An interesting observation that the kinetic rates appear to obey the Arrhenius rate law allows us to predict the dynamics of alanine dipeptide at 300 K at the microsecond timescales using the nanoseconds long high temperature calculations without actually performing MD simulations at 300 K. We conclude that the energy landscape of alanine dipeptide contains superbasins deeper than k_BT and determine the energy barriers associated with the moves from the Arrhenius rate expression. Similar insights regarding the energy landscape associated with folding/unfolding pathways of a deca-alanine molecule are obtained using kinetic rates calculated at different temperatures. (paper)