On Tue, Sep 13, 2016 at 12:19 PM, Jason Swails <jason.swails.gmail.com>
wrote:
> Actually I don't recall exposing those variables through the C API. So
> you'd need to fix *that* before even thinking about addressing the Python
> API.
>
> However, for exactly the same input (same box size, etc.), you should get
> exactly the same default PME parameter assignments. Can you run the sander
> API tests and see if those pass? Those test cases reproduce regular sander
> results exactly, if I recall correctly (well, out to the 4 decimal places
> reported in the mdout file).
>
>
> I'm not particularly concerned about the 0.0004 kcal/mol disagreement here
> in one term, though. Especially for single-point energies. But it
> *should* be identical out to 4 decimal places, so it would be interesting
> to see why it's not...
>
>
I've checked Niel's mdin and found that he set ntc and ntf to 2 while
pysander use 1
.Niel: try ntc=1, ntf=1 in your mdin: This should work (just tried).
Hai
> All the best,
> Jason
>
> On Tue, Sep 13, 2016 at 12:06 PM, Hai Nguyen <nhai.qn.gmail.com> wrote:
>
> > On Tue, Sep 13, 2016 at 11:10 AM, Niel Henriksen <shireham.gmail.com>
> > wrote:
> >
> > > Hi Dan,
> > >
> > > Can I set those through the pysander interface?
> > >
> > >
> > I think no. Those are some options:
> > https://github.com/hainm/amber_things/issues/21
> >
> > you need to change pysander code for that:
> > https://github.com/Amber-MD/pysander/blob/master/sander/
> > src/pysandermodule.c
> >
> > Hai
> >
> >
> > > --Niel
> > >
> > > On Tue, Sep 13, 2016 at 7:53 AM, Daniel Roe <daniel.r.roe.gmail.com>
> > > wrote:
> > >
> > > > Hi Niel,
> > > >
> > > > Did you explicitly set PME parameters in your input (nfft1 etc)? If
> > > > you make sure those are explicitly set you should get better
> > > > agreement.
> > > >
> > > > -Dan
> > > >
> > > > On Tue, Sep 13, 2016 at 10:34 AM, Niel Henriksen <shireham.gmail.com
> >
> > > > wrote:
> > > > > Hi Hai and Jason,
> > > > >
> > > > > Thanks for the troubleshooting advice. I tried both suggestions.
> > > Rather
> > > > > than read in the trajectory, via imin=5, I just did a single step
> > > energy
> > > > > calculation on a netcdf restart file.
> > > > >
> > > > > Sander EEL: -8225.6990
> > > > > pt.energy_decomposition elec: -8225.7691
> > > > > sander.energy_forces elec: -8225.6986
> > > > >
> > > > > So going through pysander directly is indeed a closer match. Any
> > > concern
> > > > > about the remaining difference? I can send you the files directly
> if
> > > you
> > > > > would like.
> > > > >
> > > > > Thanks,
> > > > > --Niel
> > > > > _______________________________________________
> > > > > AMBER mailing list
> > > > > AMBER.ambermd.org
> > > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > >
> > > >
> > > >
> > > > --
> > > > -------------------------
> > > > Daniel R. Roe
> > > > Laboratory of Computational Biology
> > > > National Institutes of Health, NHLBI
> > > > 5635 Fishers Ln, Rm T900
> > > > Rockville MD, 20852
> > > > https://www.lobos.nih.gov/lcb
> > > >
> > > > _______________________________________________
> > > > AMBER mailing list
> > > > AMBER.ambermd.org
> > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > >
> > > _______________________________________________
> > > AMBER mailing list
> > > AMBER.ambermd.org
> > > http://lists.ambermd.org/mailman/listinfo/amber
> > >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
>
> --
> Jason M. Swails
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Sep 13 2016 - 10:00:02 PDT
