Brian Stephenson wrote:
I have played around with tau_p, even increasing it to very high values
such as 100
and still get the pressure scaling problems. I've also energy minimized
many times,
to no avail. For this particular system I've been thinking there isn't
anything very
unphysical going on ... for some of the smaller systems I've simulated
after managing
to get the system to simulate for 100 ps or so it becomes much more
stable. I think
I simply need to get some equilibration behind me for these larger
systems as well.
If I simulate under NVT conditions I get seg faults at first but again
after getting some
simulation time behind it I believe the system could become stable. The
problem is
getting there. Simulations of the dendrimer in vacuum are stable, so it
shouldn't
be a problem in the definition of the dendrimer topology.
Btw, I've tried various tau_t and tau_p values, simulating with and
without bond constraints,
changing the simulation temperature, changing the ratio of solute (a
dendrimer) to solvent,
changing the simulation geometry (cubic box vs. dodecahedron),
decreasing the timestep,
etc. Actually the only thing that has (kind of) worked has been running
with a different
(older) version of GROMACS, which seems a bit more stable during initial
equilibration.
I'm using PME electrostatics, but the system is nonionic so that
shouldn't matter too much.
The FF is ffgmx2, and the water model is SPCE or SPC (both give the same
problems).
It really shouldn't be that hard, so I suspect that there might be other
problems than just the systems. A good way of testing is to simplify the
setup. First small systems with cutoff, then large systems with cutoff
then with PME. Also you start without pressure scaling, i.e. relax first
at NVT and when energies stabilize add pressure scaling. Same with pos res.
1. NVT with posres on protein and cut-off
2. NVT with posres on protein and PME
3. NPT with posres on protein and PME
4. NPT without posres and with PME
As you probably know the ffgmx2 field is not recommended, we usually
suggest GROMOS96 for united atom and OPLS/AA for all-atom simulations.
But these pressure scaling problems keep maxing out my harddrive, when I
would
rather that the simulation would just crash or at least not continually
write the
error message to file. Any ideas on either getting around this period
of instability
or turning off the pressure scaling errors?
Thanks,
-Brian
Message: 4
Date: Sat, 25 Feb 2006 09:48:00 +0100
From: David van der Spoel <[EMAIL PROTECTED]>
Subject: Re: [gmx-users] pressure scaling and hard drive space
To: Discussion list for GROMACS users <gmx-users@gromacs.org>
Message-ID: <[EMAIL PROTECTED]>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Brian Stephenson wrote:
> Anyone know how to turn off those pesky "pressure scaling is greater
> than 1%" error messages? Instead of crashing, my simulations keep
> running with errors and I keep running out of hard drive space when the
> simulation keeps sending out the warnings over and over again. I'm
> using mdrun with the -v flag, and maybe not using the -v would get rid
> of them (?) but I need to know how much time is left for my jobs.
if you get these warnings often your simulation is really behaving
unphysically because something is wrong.
Are you sure your simulations are OK?
If you scale temperature or pressure too fast the whole system can
become unstable.
>
> Thanks,
> Brian
Message: 1
Date: Sat, 25 Feb 2006 08:34:05 -0300
From: [EMAIL PROTECTED]
Subject: Re: [gmx-users] pressure scaling and hard drive space
To: gmx-users@gromacs.org
Message-ID: <[EMAIL PROTECTED]>
Content-Type: text/plain; charset=ISO-8859-1; format="flowed"
Hi Brian,
if you are using Berendsen pressure coupling (and your system is in the
equilibration phase) try increasing tau-p.
Regards.
Pedro.
> Anyone know how to turn off those pesky "pressure scaling is greater
> than 1%" error messages? Instead of crashing, my simulations keep
> running with errors and I keep running out of hard drive space when
> the simulation keeps sending out the warnings over and over again.
> I'm using mdrun with the -v flag, and maybe not using the -v would
> get rid of them (?) but I need to know how much time is left for my
> jobs.
>
> Thanks,
> Brian
>
_______________________________________________
gmx-users mailing list gmx-users@gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to [EMAIL PROTECTED]
Can't post? Read http://www.gromacs.org/mailing_lists/users.php
--
David.
________________________________________________________________________
David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
[EMAIL PROTECTED] [EMAIL PROTECTED] http://folding.bmc.uu.se
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
_______________________________________________
gmx-users mailing list gmx-users@gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to [EMAIL PROTECTED]
Can't post? Read http://www.gromacs.org/mailing_lists/users.php