[gmx-users] GROMOS 45A3 ff

[Pedro Alexandre Lapido Loureiro]

> 
> After some MD using ffG43a1 in a system composed of protein+solvent, I want
> to extend my work to bilayers (protein+solvent+bilayer).
> 
> I know that Dr. van Gunsteren has a new GROMOS 45a3 forcefield.  Has anyone
> done some tests with this f.f. under gmx?
> 
> I think Pedro A. L. Loureiro asked some questions about implementation of
> this f.f.. How's the work Pedro?
> 
> Or should I stay with the ffG43a2x, and hack the parameters of the lipid
> molecule from Dr. Tieleman's lipid.itp, as sugested by Erik (Mon,10 Feb
> 2003)?
> I want to maintain a certain consistency between the runs and the
> forcefield
> used, in order to compare them.
> I'm a bit confused in which f.f. to use, even after reading some stuff from
> the mailing list.
> Any comments?
> 

Olá Nuno!


This new forcefield is described in Eur. Biophys. J. (2003) 32: 67-77.
In short, it replaces the atom type of the phospholipid carbonyl (C=O) trying 
to correct a smaller-than-experimental area per lipid of the former 43a1 
forcefield.
However, I found it was necessary to use the atom type 45 (they used a 
geometric mean of atoms 45 and 12).
My simulations successfully reproduced experimental parameters (for instance, 
volume per lipid with an error of 1%), at least for the phospholipd DMPC.

Regards.

Pedro.

--
Pedro Alexandre Lapido Loureiro
Laboratório de Física Biológica
Instituto de Biofísica
UFRJ
Brasil