[gmx-users] Parameters

[Xavier Periole]

On Mon, 10 Nov 2008 00:45:19 +0000
  "andrea hanna" <al233997 at gmail.com> wrote:
> Dear users,
> 
> I have performed a number of simulations using the ffG43a2 ff in 3.2.1 and
> the following parameters (only a few are given):
> 
> ; RUN CONTROL PARAMETERS
> integrator               = md
> tinit                    = 0
> dt                       = 0.002
> nsteps                   = 125000
> comm-mode                = Linear
> nstcomm                  = 1
> 
> nstlist                  = 10
> ns-type                  = Grid
> pbc                      = xyz
> rlist                    = 1.0
> domain-decomposition     = no
> 
> coulombtype              = PME
> rcoulomb-switch          = 0
> rcoulomb                 = 1.0
> epsilon-r                = 1
> vdw-type                 = Cut-off
> rvdw-switch              = 0
> rvdw                     = 1.0
> 
> I recently noticed from the manual that the nstlist should be 5 with gromos
> 96 ff. I have a couple of questions.
> 1. should this be 5 with a 0.001 timestep or with a 0.002 timestep
> 2. As I have done a lot of work using these parameters I was really looking
> for some suggestions/advice as to whether I can use any of my data - I;m
> really clinging onto hope here! And as to how 'badly' this choice/mistake
> will have affected my work.
The nstlist is one issue but the most worrying is the use of a straight
cutoff at 1.0 nm for the vdW. This should be 1.4 nm and makes a big 
difference.

Note also that I am not sure this force field has not been tested for PME and
developed for Reaction-Field. You should check the original paper.

You can rerun some of your jobs with a parameter set closer to the original
and check whether your observable vary significantly or not. That is
probably the best you can do.

XAvier.
> 
> Thanks,
> Les

-----------------------------------------------------
XAvier Periole - PhD

- Molecular Dynamics Group -
Computation and NMR
University of Groningen
The Netherlands
-----------------------------------------------------