On Tue, 2013-12-03 at 18:25 +0000, Aronica, Pietro wrote:
> A premise: I'm working with an experimental protocol that requires a lot of custom residue and atom types and precise manipulation of the prmtop files. It has been described as "a prmtop-based FEP scheme" by Jason himself, and that's not far from the truth. Because of the settings of my system, it is crucial that the exclusion list is kept the way I originally make it in order to avoid instabilities and failures. Indeed, I manually edited the prmtop files to change only the bond length I needed to modify (which is what I expected parmed.py to do) and the issue disappeared.
> The command is as follows:
>
> setBond :MKN.CH :MKA.CG 340 1.526
> setOverwrite
> parmout Solv_100_0.prmtop
>
> There's a few more, varying from 10_90 to 100_0, with different values for the equilibrium distance and the force constant. In each case, the bond parameters are correctly modified, but in addition the number of excluded atoms goes from 28762 to 28755. I'm wading through the exclusion list to pinpoint the change, but it arises from the custom residue where the bond is modified. That is the only instance of the bond in the system.
> I figure it's probably my delicate prmtop manipulations that are causing this, but I don't understand why the setBond command should touch the exclusion list at all if the bond is already there and connectivity is unchanged.
I'm intrigued. My guess, actually, is that your original prmtop file is
broken (by tleap) (in tleap's defense, its job is much much harder than
ParmEd's job).
Can you try the following two ParmEd commands and see if the output is
helpful?
parmed.py -p <prmtop> << EOF
setMolecules True
setOverwrite
parmout test.parm7
EOF
Do you get any messages about a corrupt ATOMS_PER_MOLECULE section? If
so, please add the "setMolecules True" command to the following test as
well:
parmed.py -e -p <prmtop> << EOF
setOverwrite
!amber_prmtop.parm.remake_parm()
parmout test.parm7
EOF
Does this 'final' test.parm7 file (in which you did not _change_
anything) have the number of exclusions you expected or a different
number?
If you send me your prmtop file and your inpcrd file I will try to take
a look.
Thanks,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Dec 03 2013 - 11:30:02 PST
