I have recently started working with enhanced sampling methods (particularly the OPES explore variant introduced in this paper https://doi.org/10.1021/acs.jctc.2c00152) and it is my first experience in this field. I would appreciate if anyone could give me some advice.
I am trying to get the PMF for a organic molecule close to a crystal surface site. For this reason I am defining the water coordination number on that site and the distance along z-axis as collective variables. Since I am interested in the PMF on top of this particular site, I am restraining my molecule in a cylinder of radius R of height H by defining bound by harmonic walls. I have set H far enough to ensure it is in the bulk solution region. However, I have noticed the method is adding bias in the region just below the harmonic wall. I have some idea on what may be happening:
I would appreciate any inputs on this so I can understand what is happening and, if necessary, avoid oversampling/adding bias close to the harmonic wall in the bulk region.
Thank you!