[gmx-users] Fwd: FW: PME settings in free energy calculations

[Dries Van Rompaey]

Hi,

I'm using PME with the default Potential-Shift-Verlet modifier. If I'm
interpreting the manual and comments on the mailinglist correctly, the
settings for rcoulomb-switch shouldn't affect this and PME-switching
shouldn't be necessary (although there seems to have been some discussion
about this, https://gerrit.gromacs.org/#/c/2220/). Has a consensus been
reached elsewhere?

Current nonbonded settings in my .mdp files are:
; Electrostatics
coulombtype              = PME
rcoulomb                    = 1.0

; van der Waals
vdwtype                    = cutoff
vdw-modifier            = Potential-switch
rvdw-switch              = 0.9
rvdw                         = 1.0

; Apply long range dispersion corrections for Energy and Pressure
DispCorr                  = EnerPres

; Spacing for the PME/PPPM FFT grid
fourierspacing           = 0.08

; EWALD/PME/PPPM parameters
pme_order                = 6
ewald_rtol                 = 1e-06
epsilon_surface        = 0

Thanks in advance,

Dries

On 4 February 2016 at 00:56, Michael Shirts <mrshirts at gmail.com> wrote:

> Hi, Dries-
>
> Questions like this are probably best answered on the gmx-users list.
> I can't say too much for the Verlet scheme -- I know that it was
> relatively recently adapted for free energies, and there may be some
> combinations of settings that could give unanticipated results.
>
> Pretty much all of our experience about nonbonded calculations and
> free energies is collected in the following paper:
> http://pubs.acs.org/doi/abs/10.1021/ct4005068. We used the group
> cutoff scheme since that was the only one that was supported in
> GROMACS at the time.
>
> Since you haven't sent any files, it's hard to tell what is actually
> going on. The one thing that has a tendency to happen with the more
> recent update schemes is if you set a potential modifier that is a
> switch, but don't set the distance the switch starts at, then it is
> automatically set to zero.  Check the mdout.mdp to see if this is
> happening.
>
> > From: Van Rompaey Dries <Dries.VanRompaey at uantwerpen.be>
> Date: Wednesday, February 3, 2016 at 12:34 PM
> To: Michael Shirts <michael.shirts at colorado.edu>
> Subject: PME settings in free energy calculations
>
> Dear prof. Shirts,
>
> I'm currently trying to figure out the PME settings for a relative
> free binding energy simulation I'm working on. I took the parameters
> from Matteo Aldeghi's
> paper(
> http://pubs.rsc.org/en/content/articlelanding/2015/sc/c5sc02678d#!divAbstract
> )
> as a starting point (adapted for the Verlet scheme by scaling the
> fourierspacing to 0.08 and setting coulomb = rvdw = 1.0). I then tried
> to verify these settings by comparing a single point energy
> calculation with these settings and one with very long coulomb cutoffs
> as recommended on alchemistry.org.
>
> Unfortunately, I can't seem to get this quite right. I'm getting
> differences in the hundreds of kj/mol, leading me to suspect I'm doing
> something wrong. I'm calculating the energy values by extracting
> CoulombSR and Coul-recip from the energy.xvg files. I've tried
> calculating the energy with coulomb cutoffs at 3 nm and 10 nm, but
> agreement with the PME results remains rather poor. David Mobley
> mentioned you performed extensive research on this topic, and I'm
> hoping you could point me in the right direction.
>
> Thanks in advance,
>
> Dries
>
>
>
> ~~~~~~~~~~~~~~~~
> Michael Shirts
> Associate Professor
> michael.shirts at colorado.edu
> Phone: (303) 735-7860
> Office: JSCBB D317
> Department of Chemical and Biological Engineering
> University of Colorado Boulder
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>