Hello, I am planning on measuring the energy associated with diffusion of small molecules through a membrane using ABF. I have generated the initial reaction coordinate using steered MD (along the z-coordinate), but before I invest the computational resources in the next step I want to clarify something. The constraints for each window are only placed in the z axis (along the reaction coordinate), correct? Most ABF or umbrella sampling simulations I've seen have been through protein channels, so diffusion in the x-y plane is never an issue, but in this case the molecule will be free to diffuse all over in the x-y plane at each window. Should this movement be constrained, or should it be left completely un-constrained in the x-y plane? Thanks very much for any help.