Mapping potential energy surfaces

An approach which uses a coarse-grained non-Markovian dynamics to bias microscopic atomic trajectories to study the T=0 free energy surface was discussed. The accuracy of the reconstructed potential energy surface (PES) was dramatically improved by a simple postprocessing procedure with minor computational overhead. It was found that the PES in coarse-grained space was mapped by performing a constrained minimization every 10fs of an MD run. Analysis shows that the approach was illustrated by conducting conformational analysis on a small organic molecule, demonstrating its superiority over traditional unbiased approaches in sampling PES in coarse-grained space.