GROMACS version: 2021.5
GROMACS modification: No
Here post your question
I am performing unbiased MD to study interactions of a metal cluster (hundreds of metal atoms without any interatomic bonds) with proteins. I would like that the spherical shape of the clustered atoms remains intact during interactions as observed in our experiments. I have tried position restraints to fix the position of atoms in the metal cluster while keeping the protein unrestrained but this does not lead to expected interactions. Hence, I would like that the metal cluster exist as a free floating sphere in TIP3P water but without loosing its characteristic shape. In other words, the relative positions of atoms within the cluster should remain fixed while the cluster, as a whole, moves in the solvent (and interacts with the protein). Is there a way to do this?
type or paste code here
Maybe flat-bottomed position restraints is closer to what you look for, but it does not allow the metal particle to float around. see Restraints — GROMACS 2022 documentation
Distance restrains using simple harmonic restraints Restraints — GROMACS 2022 documentation will allow to fix the distance (not distance) between the metal atoms (but not necessary the spherical shape)
I hope it helps
Thanks for the direction. Will distance restraints allow the metal cluster to move around in water as a nearly-spherical object?
Alternatively, can constraints do the trick? If so, how can I generate a topology file for all the constrained atoms?
distance restraints will fix the distance between the metal atoms, do not fix the position of the atoms. There is one type of distance restraints that fix the distance between two atoms using harmonic potential. These types of restraint are defined in the topology (as bond type 6, see more details in the manual). This allow you to set a network of spring between the metal atoms. You might be able to tune in a way that it gives a sort of spherical shape, depending on how are defined the interaction of the metal atoms in your force field.
I do not fully understand your second point.