On Wed, Jun 19, 2013 at 10:18 PM, Fabrício Bracht <fabracht1.gmail.com>wrote:
> Hello. Could this be because this topology has a zinc atom and aminoacid
> side chains attacked to it. The parameters to all these atoms were
> generated using the mtk++ procedure. I have also changed the LJ parameters
> to the water molecule attached to the zinc atom, since the standard
> parameters caused the hydrogen atoms to collapse. I ran parmed.py check
> validity and I got a warning :
>
> Determining validity of prmtop
> /home/fabricio/amber12/bin/chemistry/amber/readparm.py:1046:
> MoleculeWarning: Molecule atoms are not contiguous! I am attempting to fix
> this, but it may take a while.
> MoleculeWarning)
> 1 total warnings
>
This is the warning I described earlier, and is the reason cpptraj is
failing to work with your files. During the simulation, the only issues
this warning may cause are the pressure scaling in NPT simulations. If
your density equilibrated as expected and everything seems reasonable, I
expect you are safe (although it may be worthwhile to run a short
simulation at NPT to compare results with the 'bad' topology file).
However, as mentioned, you will need to fix your topology file (and NetCDF
trajectory) in order for cpptraj to process your files (and it's necessary
for a number of cpptraj or ptraj actions to process your frames correctly,
like imaging). To this end, I suggest reading through my last response for
a solution to fixing your trajectories (as long as they are NetCDF) and
your topology file.
>
> NonUniversalWarning: Modified off-diagonal LJ parameters detected! Will
> probably
> not work with pmemd.cuda or mgdx!
>
Hmm... I wonder why this warning was triggered. Unless you used
"changeLJPair" in ParmEd (or somehow otherwise manually changed the
LENNARD_JONES_ACOEF or BCOEF elements of the topology file, this should not
be an issue. Can you email me your prmtop off-list so I can see why this
error is being triggered?
Thanks!
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 Wed Jun 19 2013 - 20:00:04 PDT
