I have now the same issue with small molecule ligand docked into zinc containing atom.
TER record are ok and PDB was not generated by tleap.
APPROXIMATING switch and d/dx switch using CUBIC SPLINE INTERPOLATION
using 5000.0 points per unit in tabled values
TESTING RELATIVE ERROR over r ranging from 0.0 to cutoff
| CHECK switch(x): max rel err = 0.2738E-14 at 2.422500
| CHECK d/dx switch(x): max rel err = 0.8314E-11 at 2.736960
---------------------------------------------------
| Local SIZE OF NONBOND LIST = 524773
| TOTAL SIZE OF NONBOND LIST = 3783691
partition error in shake on processor 0
this processor has atoms 1 through 3310
atom 3310 is within this range
atom 3311 is not within this range !
ATOM 3310 C6 RES 211 12.953 -7.656 8.135 1.00 0.00
ATOM 3311 H5 RES 211 12.695 -7.553 9.185 1.00 0.00
ATOM 3312 C4 RES 211 13.722 -8.746 7.726 1.00 0.00
Answers in mailing list were not helpful until so far as this is not water and TER records look ok and file was not from tleap. Maybe file of ligand used for antechamber with all hydrogens was, I ll check that if that can be relevant.
Best regards,
Andrew
02.11.2011, 09:36, "Jason Swails" <jason.swails.gmail.com>:
> I've seen tleap do very weird things determining the ATOMS_PER_MOLECULE and
> SOLVENT_POINTERS section of the topology file, so I KNOW there's a bug
> there for some corner cases (that are becoming more common). �Ross can
> attest to this as well -- it has caused weird segfaults in pmemd and sander
> with strange systems that we've both looked into.
>
> That's why I added a function in my prmtop editor (parmed) to reset this
> section completely based on the actual bonded atom network. �Obviously not
> as ideal as fixing it in leap itself, but it's a quick way of testing
> whether or not this will fix the issue...
>
> (The parmed command is "setMolecules solute_ions=True" if you were
> interested in trying this to see if it fixes the issue)
>
> All the best,
> Jason
>
> On Tue, Nov 1, 2011 at 9:48 PM, case <case.biomaps.rutgers.edu> wrote:
>
>> �On Tue, Nov 01, 2011, Breuer, Marian wrote:
>>>> �I'm currently trying to carry out MD simulations with sander (MPI
>>>> �parallel version, AMBER11) on a protein system using SHAKE bond
>>>> �constraints for bonds containing hydrogen (sander option ntc=2). These
>>>> �work with up to 14 processors, but with any number of cores higher than
>>>> �that the simulations abort in the first step with the error message:
>>>>
>>>> ��partition error in shake on processor �����������0
>>>> ��this processor has atoms �����������1 �through ��������8265
>>>> ��atom ��������8265 is within this range
>>>> ��atom ��������8266 is not within this range !
>> �OK: your prmtop file has bad information in it, so the "problem" is not
>> �with
>> �sander, but with how the prmtop file itself was generated.
>>
>> �Specifically, the ATOMS_PER_MOLECULE field in the prmtop has this:
>>
>> �%FLAG ATOMS_PER_MOLECULE
>> �%FORMAT(10I8)
>> ����8265 ����848 ������1 ������1 ������1 ������1 ������1 ������1 ������1
>>
>> �...etc., which is the reason that sander thinks that atoms 8265 and 8266
>> �are
>> �in different molecules (and hence can be assigned to different processors).
>>
>> �You seem to have a system with 1 "HER" group, 9 heme groups, a protein,
>> �and a pile of sodiums, chlorides and water. �It looks like at some point,
>> �the HER and HEM residues had their coordinates after the regular protein
>> �coordinates, but that at some point, these HER/HEM atoms got placed before
>> �the protein ones. �Also, most likely, each HEM group should be its own
>> �molecule (which is later cross=linked to the protein somehow. �This means
>> �that in the input pdb file, there should have been a TER after the protein
>> �atoms and after each HEM/HER residue. �Then (once all hydrogens are added)
>> �the first 8265 atoms would be protein (amino acid atoms); there should
>> �then be 10 HER/HEM residues. �The ions and water would then be added by
>> �LEaP.
>>
>> �My wild guess is that either your input pdb file didn't have TER cards in
>> �all
>> �the required places, or that you used the savePdb command at some point,
>> �then
>> �re-read that back into LEaP(?) It may well be a bug in LEaP and not
>> �operator
>> �error, but I can't tell.
>>
>> �But this is about as far as I can go...if seeing this description of the
>> �problem doesn't help, we would need to know the exact commands you used to
>> �create the prmtop file.
>>
>> �...good luck...dac
>>
>> �_______________________________________________
>> �AMBER mailing list
>> �AMBER.ambermd.org
>> �http://lists.ambermd.org/mailman/listinfo/amber
> --
> 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
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Nov 26 2011 - 01:00:03 PST
