I would humbly add observations from my group.
We do perform at least minimally systematic studies on most of the systems
that we simulate and we find that 18 Angstroms is usually where things begin
to behave well. By 'things', I mean the density, counter-ion diffusion
rates, and so forth. By 'behave well', I mean that their converged values
become reasonably independent of the buffer value.
We have not done anything with membrane systems in my group, so you probably
should do at least a few tests of your own.
Cheers,
Brent
On Tue, Aug 24, 2010 at 12:40 PM, case <case.biomaps.rutgers.edu> wrote:
> On Tue, Aug 24, 2010, colvin wrote:
>
> > 1. Is there any rule or guide in setting the distance value for
> > solvateBox?
>
> In principle, the bigger the box, the smaller will be the periodic
> artifacts.
> For "garden-variety" globular, soluble proteins in water, I often use 12
> Ang.
> as the buffer. But this is not based on any systematic study. You seem to
> have a protein-membrane system, but I don't know if it has any
> two-dimensional
> periodicity along the membrane plane or not.
>
> One general suggestion is to find a paper that has done simulations like
> the
> ones you are considering, and use their setup as a starting point. Better,
> do a more thorough literature search as see what the current state of the
> art
> is.
>
> ....dac
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
_______________________________________________
Brent P. Krueger.....................phone: 616 395 7629
Associate Professor................fax: 616 395 7118
Hope College..........................Schaap Hall 2120
Department of Chemistry
Holland, MI 49423
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Aug 24 2010 - 10:30:04 PDT
