I have two zinc coordinations in my globular protein, using Amberssff in Gromacs. Partial charges have been calculated for both using PyRED. One zinc remains in place using a non-bonded model, mainly due to the fact that it is buried in the protein and the only electronegative neighbouring atoms are the coordinating nitrogens on its four histidines.
The second zinc complex uses water as a fourth donor. Thus I am trying to use covalent bonds to keep the zinc in place with its three histidines. I have partial charges for the histidines, however, when I have tried to refit the unknown force field parameters for the Zn-N bonds using PyRED I get nothing because they don't exist in the Amber bonded parameter data set. It turns out I need to supply a frcmod.user file to make up for everything missing in the frcmod.unknown (all my Zinc related parameters). Pretty understandable.
I was able to find Amber Zn-N bond distance force constants and bond distance equilibrium values, and even so for N-Zn-N angles. Unfortunately the dihedral angles have been a little trickier to find. My first questions involves the nature and number of dihedral angles. With zinc surrounded by three histidines (plus one water), donating nitrogen to the first coordination sphere, what particular dihedral angles should I include in an MD forcefield?
Zn - N - C - C
Zn - N - C - N
N - Zn - N - C (db bonds to other carbon on ring)
N - Zn - N - C (sg bonds to the second nitrogen on the ring)
My second question. Is it possible to run Gromacs and does it make sense to implement only [ bonds ] entries for Zn-N* entries? Or do I need to also have angles and dihedrals?
Thanks
Anthony
The first thing to realize is that dihedrals are basically fudge factors for inaccuracies in nonbonded parameters; you don't always need them. You often do, but in a case like this, where a transition metal is already being assigned a fixed charge, as long as you keep the geometry reasonably sane with bonds and angles, I don't know that dihedrals are needed. If you identify some specific deficiency, then you may have to parametrize it (likely the only way to do that here is to dive into QM).
Often, people just use simple distance restraints to keep transition metals in place. This seems to work reasonably well, but I suppose if you have decent bond and angle parameters, that's all you should need.
Thanks Justin, that was of huge help. I now have a fully parameterised force field. Fingers crossed that I get it through grompp and an energy min stage!
- Badges
- Science topic