[gmx-users] special molecule

[Mark Abraham]

Dongsheng Zhang wrote:
> Sear Mark,
> 
> Since PRODRG server only generates a itp file by using gromos type force
> field, do I have to use gromos force field for my whole protein-ligand
> system if my ligand force field parameters are got from PRODRG? In
> another word, can I mix two different force field (one for protein, one
> for ligand) in one system? Thanks!

It's technically possible, if fiddly, but I'd strongly advise against 
it. Force fields are mathematical constructs that are optimized to 
produce reasonable model physics as an interacting unit. There is no 
reason to suppose that a bond stretching parameter is a measure of bond 
strength, for example, because the other parameters in this force field 
might have some quirk that this parameter is compensating for because of 
the optimization procedure. A force field is merely a point in a 
high-dimensional space which when used the right way has been shown to 
produce a useful physical model. Thus there is no reason to suppose that 
a few of those high-dimensional space coordinates plus some others from 
some other force field will combine to produce a useful physical model.

Mark