[gmx-users] changing partial charge/atom type upon covalent bonding

[Leandro Bortot]

Dear users,

     I need to simulate a protein which is covalently bonded to some
covalent inhibitors using GROMACS 5.0.4 and the AMBER99SB-ILDN force field.
     Thanks to the user discussion list I successfully used specbond.dat
and suitable modifications in a local copy of the forcefield to make the
relevant bonds between the protein and the inhibitors. However, I have a
specific question about how account for the effect of the new bonds on the
pre-existing parameters.
     As an example, let's take an ester bond with a Glutamate side chain. I
added it to specbond.dat. The GLU side chain partial charges and atom types
should change as a result of the covalent bond. Hence, I created the GLX
residue so I can change its partial charges and atom types. But how should
I derive them?

     My idea is to make a GAFF/Antechamber/acpype parametrization of the
ligand bound to a glutamate residue.
     By doing this, I will be able to get the charges and atomtypes for the
bonded residue and add them to the forcefield as the "post bond" GLX
residue. Additionally, I will have better parameters for describing this
new bond itself.
     Is this approach correct? Any suggestions?

Thank you for your attention,
Leandro.