Hi, I’m trying to follow a paper where they’ve carried out simulations using LAMMPS, where they introduce Gaussian potentials with a specified width and depth, at fixed positions. Is there a way to implement something like this in GROMACS? In essence this is required to hold the atoms in place. I know you can use position restraints in GROMACS, but that gives a positive restraint potential, whereas I’m trying to implement a negative (attractive) potential between this made up potential well and an atom.
Do you want to do this for all atoms or just a few? If it’s just a few, you can add an atom to the molecule type which has no interactions apart from a tabulated bonded potential with the atom of interest. Then you can freeze the extra atom and provide the Gaussian potential through a table.
I’ve been giving this quite a bit more thought, and I think I reached a similar conclusion.
I was thinking of adding virtual sites (with 0 mass) on top of the atoms, and restraining these sites to the initial configuration of the system. Then I would introduce an attractive non-bonded potential between the virtual site and its respective atom using a tabulated potential which is tabled as (half) of an inverted normal distribution, with the (negative) minimum of the normal distribution being at x = 0, and the maximum potential being = 0.
In my case the atoms form part of molecules, and I want those atoms to still interact using the normal bonded potentials, and non-bonded potentials with other atoms in other molecules.
The problem I’m facing now is that I’m using GROMACS 2021, and tabulated potentials don’t seem to work beyond GROMACS 2019. So I’ll either have to downgrade to GROMACS 2019, or find another solution, although I can’t really think of any other solution.
I also need to figure out how to reference the tabulated potentials. I understand the table, and the form of the potential function, but referencing everything in the topology, and the mdp file, is another issue.