[gmx-users] How to include bonded parameters in ffgmxbon.itp?

[Erik Lindahl]

Alexandre Suman de Araujo wrote:
> Hi GMXer
> 
> I'm trying to include parameters to acetonitrile in ffgmxnb.itp and
> ffgmxbon.itp. The Non-bonded parameters I can include using LJ parameters
> and charges found in some papers. My doubt is how to find(or calculate)
> bonded parameters to include in ffgmxbon.itp?
> 
> Thank's and waiting answers
> 

Hi Alexandre,

First - bonded parameters in most forcefields are not that well 
determined; a lot of the are actually more or less intelligent guesses.

If you constrain the bonds, the actual force constant won't matter, only 
  the bondlength, so that removes one parameter. The bond distance between
two atom types probably don't differ that much between force fields; the
important thing is to browse the force fields and find atoms with 
similar environments.

When it comes to the dihedrals, it's important that you choose a 
potential that is consistent with your treatment of 1,4 interactions, so 
if you use Gromos96 nonbonded parameters I would recommend the Gromos96 
dihedrals too (instead of e.g. OPLS).

Cheers,

Erik