TRAMPLE: maximum likelihood estimation of protein unfolding rates from all-atom temperature ramp simulations

Efficiently estimating the kinetic or thermodynamic stability of single-chain proteins or protein-protein complexes is both notoriously difficult as well as important to computational protein and peptide design. Here, we present a new all-atom molecular dynamics-based enhanced sampling method, Temperature Ramp Maximum Likelihood Estimation (TRAMPLE), in which we estimate the kinetic stability by explicitly fitting the terms in the Eyring equation using protein unfolding trajectories generated under fixed temperature ramping schedules.