I have an enzyme and a protein ligand bound. In the active site of the enzyme is a ZN2+ ion. I am using the latest version of gromacs with the Amber99SBFF. After carefully settling my solvated system down with a serious of energy minimisation and short MD simulations with position restraints on the backbone atom and zinc atoms, an unrestrained system results in the zinc working its way free and vanishing into solvent. The forcefield is parameterised for zinc, and I have changed all short range cut offs, and NPT parameters to match that of an out of box Amber simulation, as I am able to keep the zinc ion in the active site using the Amber code.
Would it be advisable to parameterise the forcefield myself for the zinc active binding site using ab initio software (e.g., DFT, through Gaussian). I am getting the impression that the specific interactions between ZN and the binding site sidechains are active-site dependent across different enzymes.
Hi Stephane,
I changed the short range cut offs in the gromacs .mdp files to reflect those used in the Amber simulations (0.8 nm, not picked or ran by myself). PBC are ok, and the AmberFF is the latest provided by the latest gromacs distribution (so it is not something that I have either changed or implemented). I have compared the parameters of the Amber code/AmberFF and the Gromacs code-AmberFF, they are the same. I think the difference in the Amber code/AmberFF trajectory and the Gromacs code-AmberFF trajectory are by chance alone, and if I was to run the Amber code for longer the Zn+ would leave the active site eventually.
The Zn+ is ligated to three histidines (as per the crystal structure and accompanying publication from the protein data bank) in the enzyme active site. So I have taken the coordinates of the three histidines and the zinc ion and I am performing some DFT calculations over a few different degree of basis sets. From looking at the first complete DFT calculation, the partial charges are very different to those of the three histidines generated with the amber force field.
This is my first attempt at parameterising a FF for a specific transition metal-ligand in mind.
Many thanks
Brilliant! Thanks for supplying those. I will go through them carefully.
Thanks
- Badges
- Science topic