Efficient acceleration of the convergence of the Minimum Free Energy Path via path-planning generated initial guess

We demonstrate how the combination of a shifted clustering algorithm and a fast-marching- based algorithm is able to generate good approximations of the Minimum Free Energy Path (MFEP) if a Free Energy Landscape (FEL) is given. Then, we show that using this kind of approximation as the MFEP’s first guess and the string method for further refinement (also called the FMT-string combined approach) cuts down on the number of iterations needed for the MFEP to converge by a large amount. This saves a lot of time compared to using a linear interpolation as the first guess. Such a method provides an efficient alternative to the growing string method for obtaining a good initial guess of the MFEP.