Hi Daniel:
The title indicates that I was running AMBER FF. Perhaps I should have
specified 99SB. I was simply asking whether the flags I used for TIP3 water
are correct when using NAMD-CUDA as engine. I wanted to freeze water O-H
bonds only. As the simulation proved slow, I was wondering whether my aim
was accomplished by the flags used.
I have my reasons to use that engine and AMBER FF. The first, because, once
the system is equilibrated, I want to run a particular biased MD that is
not available with the AMBER engine, either sander or pmemd. The second,
because transition metal prosthetic groups, as in my protein, can be rather
smoothly parameterized under 99SB.
francesco
On Tue, May 28, 2013 at 11:08 AM, Daniel Sindhikara <sindhikara.gmail.com>wrote:
> Hi,
> This is the AMBER mailing list. Perhaps you were looking for the NAMD
> mailing list?
> http://www.ks.uiuc.edu/Research/namd/mailing_list/
>
>
> On Tue, May 28, 2013 at 4:43 PM, Francesco Pietra <chiendarret.gmail.com
> >wrote:
>
> > Hello:
> > in NAMD-CUDA conf file, setting for explicit TIP3 water
> >
> > ## Integrator
> > timestep 1.0 # 1 fs/step
> > # rigidBonds all #with H; needed for 2fs/step
> > # rigidIteractions 100 #shake max iteractions
> > rigidBonds water
> > rigidTolerance 0.000001
> > # useSettle on #faster than shake for water
> >
> >
> > leads to
> > FATAL ERROR: UNABLE TO FIND BOND PARAMETERS FOR HX HX (ATOMS 5635 5633)
> > ------------- Processor 0 Exiting: Called CmiAbort ------------
> > Reason: FATAL ERROR: UNABLE TO FIND BOND PARAMETERS FOR HX HX (ATOMS 5635
> > 5633)
> >
> >
> > I remember - unfortunately vaguely - that there is an issue related to
> how
> > NAMD treats TIP3 water. Commenting out rigidBonds for water, as with
> >
> > ## Integrator
> > timestep 1.0
> > # rigidBonds all #with H; needed for 2fs/step
> > # rigidIteractions 100 #shake max iteractions
> > # rigidBonds water # if 1fs/step
> > # rigidTolerance 1e-08
> > # useSettle on #faster than shake for water
> >
> > the simulation goes on without errors,. Though, it is much slower than by
> > treating with CHARMM/NAMD (with flag rigidRonds water) a system of
> similar
> > size (200,000 atoms). I suspect that the above solution is affected by
> flag
> > errors. I want to stick to ts = 1 fs.
> >
> > Thanks for advice
> >
> > francesco pietra
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
>
> --
> Dr. Daniel J. Sindhikara <http://www.dansindhikara.com/Information.html>
> Ritsumeikan University <http://www.ritsumei.ac.jp/eng/>
> sindhikara.gmail.com <http://www.dansindhikara.com>
> --
> _______________________________________________
> 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 May 28 2013 - 08:00:02 PDT
