AMBER Archive (2007)

Subject: Re: AMBER: Problem with anisotropic pressure scaling

From: Robert Duke (rduke_at_email.unc.edu)
Date: Fri Mar 16 2007 - 09:46:58 CST

Linda -
Hmmm. I presume that you are actually using pmemd 3.1 here? There was no
release of pmemd with the actual amber 7 distribution; instead we released
pmemd 3.0 thru pmemd 3.1 as a tarball on the website in the amber 7
timeframe. The pmemd 3.x releases were all based on amber 6 sander code. I
think in that codebase ntp .eq. 2 was not supported. In any case, it is
clearly not supported in pmemd 3.x (you should actually be getting a message
to that effect in the mdout from the master - try a single processor run to
get away from the mpi-induced noise and you may see it). Note, however,
that ntp .eq. 2 IS supported in pmemd 9 (and probably was in pmemd 8 too).
PMEMD 9 is also a lot faster than pmemd 3.x (so maybe consider upgrading?).
Best Regards - Bob Duke (the pmemd guy)

----- Original Message -----
From: "Linda Prengaman" <lindap_at_cmu.edu>
To: <amber_at_scripps.edu>
Sent: Friday, March 16, 2007 10:23 AM
Subject: AMBER: Problem with anisotropic pressure scaling

> Dear Fellow AMBER Users:
>
> I am trying to do a constant pressure MD run on a DMPC lipid bilayer. I
> would like to set ntp=2, as that seems more appropriate for membrane
> simulations. However, when I try to run the simulation, I get the
> following error message in the output file:
>
> Unit 5 Error on OPEN: lipid_only_constP_NTP.in
>
> This is a pmemd run (amber7).
> I also get the following error in the nohup.out file (I am using no
> hookup)
>
>
> Unit 5 Error on OPEN: lipid_only_constP_NTP.in
> [0] MPI Abort by user Aborting program !
> [0] Aborting program!
> p0_11748: p4_error: : 1
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
> /usr/local/mpich-1.2.4/bin/mpirun: line 1: 11748 Broken pipe
>
> The exact same job with ntp=1 runs with no error message.
>
> Here is the input file in question:
>
> &cntrl
> nstlim=100000,
> imin=0,nscm=1000,
> nmropt=0,
> ntx=1,irest=0,ntrx=1,
> ntxo=1,ntpr=500,ntwr=500,ntwx=500,ntwv=0,ntwe=0,iwrap=1,
> ntf=2,ntc=2,dt=0.002,
> scnb=8.0,scee=1.2,
> cut=12,ntb=2,
> ntp=2,taup=0.2,pres0=1.0,plevel=2,
> ntt=1,temp0=100,tempi=3,ig=3455476,tautp=1.0
> &end
> Lipid Constraints
> 100.0
> ATOM 1 5888
> END
> END
>
> The box is orthogonal, so I don't understand what other problem there
> could be. Please help.
>
> Thank you,
> Linda Prengaman
> Department of Chemistry
> Carnegie Mellon University
>
> -----------------------------------------------------------------------
> 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