On Tue, Feb 12, 2013 at 1:23 AM, <psu4.uic.edu> wrote:
> Dear Amber,
>
> Amber 12 was built with Intel/11.1.
>
> In the _MMPBSA_gb.mdin, it says:
> Error opening unit 30: File "_MMPBSA_dummycomplex.inpcrd" is missing or
> unreadable
>
This looks like the error that was fixed by bugfix.8 in AmberTools 12. But
you say you saw this in _MMPBSA_gb.mdin? That should not be possible...
Do you mean _MMPBSA_complex_gb.mdout.0?
> Interestingly, it is the same error message as this message:
>
> http://archive.ambermd.org/201205/0651.html
>
> We have tried Jason's suggestion, which is *"add a space before the &cntrl
> and the / character at the end of that file (you can use -make-mdins to
> generate the MDIN files, change them, and then use the same command-line
> with -use-mdins instead of -make-mdins to use your modified
> copies).".* However, we add a space before the &general instead of the
> &cntrl in the
> mmpbsa.in file. The same error message occurs.
>
Putting a space in the MMPBSA.py input file has no effect, i was talking
about adding a space to the &cntrl namelist of _MMPBSA_gb.mdin file (but it
must have been written for sander). I don't think this should have been
necessary...
> We try to use Bill's suggestion
> here<http://archive.ambermd.org/201205/0651.html>
>
> "I believe that you should be able to remove 'radiopt=0' from the &pb
> section all of the mdin files created by MMPBSA.py (one of the complex,
> receptor, and ligand). Afterwards, re-run the calculation with the
> -use-mdins flag and you shouldn't see this error anymore. "
>
> So we edit _MMPBSA_gb.mdin like the following and try again, the same
> error message still occurs.
>
> GB
> saltcon = 0.1
> gbsa = 1
> extdiel = 78.3
> surften = 0.0072
> rgbmax = 25.0
> igb = 7
>
Ugh. How is this possible, either? igb=7 doesn't exist in NAB (yet), so
MMPBSA.py should use sander (and it should do so whether or not
use_sander=1 is present in the input file). Furthermore, MMPBSA.py
shouldn't even look for _MMPBSA_dummycomplex.inpcrd if it's using NAB, so
how you got the original error message is puzzling.
At this point I'm completely lost. I'm unsure of exactly what you did,
what version you're using, what changes you've made to which files and
when, etc...
If you have made any changes to the source code, undo them. Go back to the
original input file you had:
Input file for running mmPBSA
&general
keep_files=2,use_sander=1,
/
&gb
igb=7, saltcon=0.100
/
See if you still get the &cntrl namelist not found error message in the
_MMPBSA_complex_gb.mdout.0 file. If yes, please post the _MMPBSA_gb.mdin
file here. *This* file (_MMPBSA_gb.mdin) should be changed so that there
is a leading space in front of &cntrl and /, which begin and terminate the
namelist, respectively (maybe). I can't help much more than this since I
can't actually reproduce any of the behavior you're seeing (so I can only
speculate about what is going wrong).
Also make sure the MMPBSA.py tests pass. You can go to
$AMBERHOME/AmberTools/test/mmpbsa_py and type "make test". Do any of those
tests fail?
Good luck,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Feb 12 2013 - 05:30:03 PST
