[gmx-users] Optimum slice size (-sl)

[Ashutosh Akshay Shah]

Thank you Andre and Warren!

A note about PME-
The basic principle is that a system of particles is converted into a grid
(or "mesh") of density values. The potential is then solved for this
density grid, and forces are applied to each particle based on what cell it
is in, and where in the cell it lies. - from wikipedia

This is a graph of potential across a box that I am simulating, using
different number of slices. I have used the PME method in all cases. The
minimum ion diameter is 0.25 nm. Note that my box is 12 nm in z-direction,
but only the 3 to 9 nm portion is of significance (rest is vacuum).

http://s21.postimg.org/s75ck7tdj/potentialslicescomparison.png

The graph shows that more the slice size, better the resolution, even for
PME. (no visible noise at higher slice numbers)

What should I conclude out of this? Higher the number of slices, better?
Even using PME?

Ashutosh Shah
Joint School of Nanoscience and Nanoengineering




On Sat, Feb 13, 2016 at 1:56 PM, Warren Gallin <wgallin at ualberta.ca> wrote:

> If you are using a slice and PME, then wouldn’t the slice have to be thick
> enough to avoid any ion interacting with its own image in adjacent slices?
>
> It seems to me that thickness would have to be greater than the ionic
> radius at the very least.
>
> Warren Gallin
>
> > On Feb 13, 2016, at 8:39 AM, André Farias de Moura <moura at ufscar.br>
> wrote:
> >
> > the ions may have a radius, but they are distributed continuously along
> the
> > system, so you may use slice thicknesses smaller than the ionic radius
> > you're interested in.
> >
> > if the simulation was made on some sort of lattice system (which was very
> > common back in the 1990's), then the shortest, physically meaningful
> slice
> > thickness would be the lattice width. For simulations with freely-moving
> > species, you should only mind about precision: if thickness is too small,
> > density profiles may become too noisy, if it is too large you have loose
> > details of the density profile (maxima, minima, etc), so it's a matter or
> > trial and error to find out what slice thickness suites you better.
> >
> > I hope it helps
> >
> > Andre
> >
> >
> > On Sat, Feb 13, 2016 at 11:57 AM, Ashutosh Akshay Shah <
> > ashah at aggies.ncat.edu> wrote:
> >
> >> Hello GROMACS users,
> >>
> >> What is the optimum slice size when trying to capture as much variation
> as
> >> possible?
> >>
> >> Consider a box with an electrolyte having anions and cations
> distributed in
> >> it.
> >>
> >>
> >> *Would be accurate to specify a slice size smaller than the diameter of
> the
> >> ions while calculating the charge variation across the box?*
> >> Thanks!
> >>
> >> Ashutosh Shah
> >> Joint School of Nanoscience and Nanoengineering
> >> --
> >> 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.
> >>
> >
> >
> >
> > --
> > _____________
> >
> > Prof. Dr. André Farias de Moura
> > Department of Chemistry
> > Federal University of São Carlos
> > São Carlos - Brazil
> > phone: +55-16-3351-8090
> > --
> > 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.
>
> --
> 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.
>