Test of a method for sampling the internal degrees of freedom of a flexible solute molecule based on adiabatic decoupling and temperature or force scaling

Simulation of the folding equilibrium of a polypeptide in solution is a computational challenge. Standard molecular dynamics (MD) simulations of such systems cover hundreds of nanoseconds, which is barely sufficient to obtain converged ensemble averages for properties that depend both on folded and unfolded peptide conformations. If one is not interested in dynamical properties of the solute, techniques to enhance the conformational sampling can be used to obtain the equilibrium properties more efficiently. Here the effect on particular equilibrium properties at 298?K of adiabatically decoupling the motion a β-hepta-peptide from the motion of the solvent and subsequently up-scaling its temperature or down-scaling the forces acting on it is investigated. The ensemble averages and rate of convergence are compared to those for standard MD simulations at two different temperatures and a simulation in which the temperature of the solute is increased to 340?K while keeping the solvent at 298?K. Adiabatic decoupling with a solute mass scaling factor s _m ?=?100 and a temperature scaling factor of s _T ?=?1.1 seems to slightly increase the convergence of several properties such as enthalpy of folding, NMR NOE atom–atom distances and ³J-couplings compared to a standard MD simulation at 298?K. Convergence is still slower than that observed at 340?K. The system with a temperature of 340?K for the solute and 298?K for the solvent without scaling of the mass converges fastest. Using a force scaling factor s _V ?=?0.909 perturbs the system too much and leads to a destabilization of the folded structure. The sampling efficiency and possible distortive effects on the configurational distribution of the solute degrees of freedom due to adiabatic decoupling and temperature or force scaling are also analysed for a simpler model, a dichloroethane molecule in water. It appears that an up-scaling of the mass of the solute reduces the sampling more than the subsequent up-scaling of the temperature or down-scaling of the force enhances it. This means that adiabatic decoupling the solute degrees of freedom from the solvent ones followed by an up-scaling of temperature of down-scaling of the forces does not lead to significantly enhanced sampling of the folding equilibrium.