AMBER Archive (2008)

Subject: RE: AMBER: vlimit-problems after image and restart

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Tue Apr 22 2008 - 23:45:29 CDT

Hi Anselm

My guess here would be that the imaging didn't quite work right such that
the restart file you have created has a problem in some way. This could
particularly be true if you used the trajectory file to build the new
restart file. The coordinates should be compatible (although the precision
is different) but I still think this could be a little tricky. It would be
very easy to have a minor formatting problem in there - e.g. a carriage
return in the wrong place after the coordinates and before the velocities or
possibly incorrect box information. You should also note that because amber
uses leap frog there the coordinates are out of step with the velocities.
I'm not sure how this propagates to the trajectory file but it is very
possible that the velocities you have don't exactly match the frame of
coordinates that you have.

My advice would be to go back to an earlier restart file if you have one.
Before the overflow problem occurs. I always make sure I run simulations in
small blocks, say 5ns each and keep the restart file in each case so I can
always go back if need be.

You probably also want to include iwrap=1 in the &cntrl namelist so this
problem doesn't occur again.

Good luck,
Ross

/\
\/
|\oss Walker

| Assistant Research Professor |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk | PGP Key available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.

> -----Original Message-----
> From: owner-amber_at_scripps.edu
> [mailto:owner-amber_at_scripps.edu] On Behalf Of Anselm Horn
> Sent: Tuesday, April 22, 2008 02:05
> To: amber_at_scripps.edu
> Subject: AMBER: vlimit-problems after image and restart
>
> Dear all,
>
> I recently encountered a behaviour in pmemd (Amber9), that I do not
> understand:
>
> In a protein MD simulation with explicit solvent and periodic boundary
> conditions some of the water molecules adopted such distant positions
> that a format overflow in the restart file took place. Using ptraj and
> the coordinates from the ASCII trajectory file, I imaged all solvent
> molecules back into the box; the protein then protruded
> partly over the
> water box boundaries, which I don't consider an issue (but please
> correct me, if I am wrong here).
>
> I also copied the velocity information from the original
> restart file to
> the imaged one.
>
> Restarting the pmemd calculation, however, yields "vlimit exceeded for
> step" errors printed on STDOUT for the first 4 steps, and then the
> program terminates, although the simulation had run without
> problems before.
>
> When I did an MD simulation with just a single step to obtain the
> energies, I found that the energy contributions from bonds, angles,
> dihedrals and electrostatics are very similar to those of the original
> restart structure, but vdW is five times larger, EPtot
> halved, and EKtot
> almost 10,000 times larger. Obviously, there seems something
> wrong with
> the velocities. However, I ensured that the velocity
> information in the
> imaged restart file is really identical to that of the original one.
>
> Can someone explain this to me? Where is my mistake?
> Any hints are welcome.
>
> Thanks in advance,
>
> Anselm Horn
> Bioinformatik
> Friedrich-Alexander-Universität Erlangen-Nürnberg
> Germany
>
>
>
>
>
>
> --------------------------------------------------------------
> ---------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu