Umbrella sampling is an enhanced sampling method for computing the potential of mean force along a user-defined reaction coordinate(s) that would be otherwise be hard to sample adequately with conventional MD techniques.
During umbrella sampling, an additional bias potential is used to overcome high free energy barriers. The bias is a harmonic potential that is applied along a specified reaction coordinate or collective variable (CV). Typically, simulations are run for a set of windows covering the complete CV. The bias is applied to each window to restrain the system to sample in and around the window, thus ensuring potential energy overlap between adjacent windows and enabling the energy profiles for each window to be combined to obtain a complete free energy profile.
Path collective variables (PCVs) is a path-based method where two collective variables are employed to describe a predefined path. One variable measures the progress along the putative path and the other variable measures the distance from the path. The combined use of PCV and US enhances simulations sampling and allows for the exploration of low energy paths.
Umbrella Sampling (US) was used in the following examples: