AMBER Archive (2008)

Subject: Re: AMBER: pmemd

From: Robert Duke (rduke_at_email.unc.edu)
Date: Thu Nov 20 2008 - 10:40:36 CST

Well, at a minimum we would need the mdin input and info about the system (reflected in the prmtop/inpcrd) to even begin guessing. PMEMD does a lot of what sander does, but not everything (the differences are detailed in the manual). Also, presuming the executable for pmemd is good and the lam installation is good, you should have at least gotten some output in run11.out, or in your stderr file, wherever that is going. How many cpu's are you attempting to use? How many did you use for sander.MPI?
Regards - Bob Duke
  ----- Original Message -----
  From: Jeremy Harris
  To: amber_at_scripps.edu
  Sent: Thursday, November 20, 2008 11:27 AM
  Subject: AMBER: pmemd

  Hello,

  I am simulating a protein in explicit water, and I've been using sander with lamwrapper, but want to use pmemd, as I understand it is faster. However, I've been having trouble running the jobs. I'm not sure where to start with this, so any help would be appreciated.

  My submission script is:

  #PBS -N run11
  #PBS -q s2
  #PBS -l walltime=00:15:00
  cd $PBS_O_WORKDIR
  #
  #Write usefull things to the output file
  #
  echo "Starting job $PBS_JOBID"
  echo
  echo "PBS assigned these nodes:"
  cat $PBS_NODEFILE
  echo
  #
  echo First task running on:
  uname -a
  echo at
  date
  #
  export PATH=/usr/local/amber9:$PATH
  /usr/local/bin/lamwrapper -- /usr/local/amber9/exe/pmemd -O -i run.in -o run11.out -p 1CDL_wat.prmtop -c run10.rst -r run11.rst -x run11.mdcrd

  The output file is:

  PBS assigned these nodes:
  comp00
  comp00

  First task running on:
  Linux comp00 2.6.11.4-21.10-1SCORE-smp #1 SMP Tue Jan 10 16:02:23 GMT 2006 x86_64 x86_64 x86_64 GNU/Linux
  at
  Thu Nov 20 15:53:18 GMT 2008

  LAM 7.1.1/MPI 2 C++/ROMIO - Indiana University

  No dedicated nodes so no cleanup done.

  Thanks in advance,
  Jeremy Harris

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo_at_scripps.edu