Hi David,
Thank you for the tips. I made the changes you suggested, and things seem to work (at least it's creating a much larger forcedump file with different and reasonable-looking numbers in each frame). I still have to check if the results are correct.
I didn't have to change anything in sander.F90. I changed debug.F90 (to prevent the program from exiting if imin = 5, as you suggested) and trajene.F90 (asking it to call the debug_frc subroutine before the minimization begins.) Also, I changed the I/O mode for forcedump.dat to "append" when imin=5, so it doesn't overwrite the previous force dumps.
Regarding appended files, there's the small issue that the user could accidentally append a file from the previous run. It makes the most sense for the file to be deleted at the start of the calculation, but subsequent force dumps should append to the file. Is there a standard way to do this?
Thanks,
- Lee-Ping
----- Original Message -----
From: "David A Case" <case.biomaps.rutgers.edu>
To: "AMBER Mailing List" <amber.ambermd.org>
Sent: Tuesday, May 29, 2012 5:00:39 AM
Subject: Re: [AMBER] Looping over a trajectory
On Tue, May 29, 2012, Lee-Ping Wang wrote:
> energies and forces for each conformation. I was able to calculate forces
> on a single conformation by running the command: "sander -i force.mdin -o
> force.mdout -p res.prmtop -c res.inpcrd" with the following input file:
>
> do_debugf = 1, dumpfrc = 1,
> However, when I set imin = 5 and call sander with the extra argument "-y
> equilibrate.mdcrd", it still appears to calculate forces for just one
> conformation.
The debug code automatically stops after a single point calculation. It
should be possible to change the code (e.g. at the end of the debug_frc
subroutine) to avoid this. Be sure to only use this with imin=5, and to check
that the results of multiple conformations indeed match multiple runs of a
single conformation.
[An alternate is just to modify the main sander routine to print out forces at
each step....]
...good luck....dac
_______________________________________________
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 29 2012 - 10:30:04 PDT
