I haven't tried this exactly, but I've recently started running simulations
with CHARMM36 after several years using AMBER force fields. I'll throw in
my two cents here, but am also interested in what your results will be.
I think the technically correct settings for CHARMM LJ interactions are a
FORCE switch between 10 and 12 with NO long-range correction (since the
standard correction is simply not correct in this case). This is reportedly
what people in the membrane protein community use.
Some observations/notes:
1.) This cannot be matched in AMBER (since no LJ switch is available), nor
is it the default in NAMD (where most people do membrane simulations)
2.) This causes rather significant changes in the virial behavior of water
models (try running a TIP3P water box at constant pressure with no LJ
correction, the density is considerably different)
3.) CHARMM also formally uses a different water model "TIP3Pm", with small
LJ radii on the hydrogens
4.) Supposedly running membranes at constant volume is not recommended
(maybe this is just CHARMM36?), and may even be unstable. Something to do
with the surface area fluctuations?
Brian
On Wed, May 27, 2015 at 3:50 PM, Jason Swails <jason.swails.gmail.com>
wrote:
> On Wed, May 27, 2015 at 3:22 PM, Shalini Gupta <gshalini21294.gmail.com>
> wrote:
>
> > Dear Amber user group,
> >
> > I am an undergraduate student in the Grabe and DeGrado labs at UCSF,
> and I
> > had a question concerning lipid simulations. I am attempting to closely
> > mirror the simulations carried out by Klauda and co-workers (Klauda,
> J.B.
> > et al.; J. Phys. Chem. B, 2010, 114, 7830-7843) using Amber with the
> > CHARMM36 forcefield. One concern that I have is that the barostat used in
> > the Klauda paper (Nose-Hoover Piston is used to maintain pressure) and
> the
> > Lenard-Jones switching function can't be implemented in Amber. Does
> anyone
> > have experience with such an endeavor? In discussions with Richard W.
> > Pastor, it has been clear that these parameters are particularly
> important
> > for getting bilayer properties correct.
> >
>
> Amber 14 supports a Monte Carlo barostat which should be as good as the
> Langevin piston. Richard Pastor's concern with respect to Amber barostats
> is almost certainly aimed at the Berendsen barostat, which has known
> problems in terms of producing incorrect volume distributions (the
> fluctuations are too small). The MC barostat does not suffer from this
> problem.
>
> As for the L-J switch, Amber has no such option. CHARMM-GUI spits out
> Amber-compatible input files if you generate your bilayer that way, and
> their recommended approach is to simply use a 10 to 12 A cutoff with no
> switch, and it seems to do reasonably well.
>
> HTH,
> Jason
>
> -
> -
> Jason M. Swails
> BioMaPS,
> Rutgers University
> Postdoctoral Researcher
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Brian Radak
Postdoctoral Scholar
University of Chicago
Department of Biochemistry & Molecular Biology
Gordon Center for Integrative Science, W323A
929 E. 57th St.
Chicago, IL 60637-1454
Tel: 773/834-2812
e-mail: radak.uchicago.edu
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 27 2015 - 14:30:02 PDT
