Hi Jason -
I think there is no issue with the sum of ATOMS_PER_MOLECULE being equal
to NATOM. The situation is as follows:
The .pdb file looks like the following.
======================== .pdb file ===============================
ATOM 1 HO5' DG5 1 -24.493 31.739 17.480 1.00 0.00
ATOM 2 O5' DG5 1 -25.368 31.480 17.777 1.00 0.00
.
.
ATOM 724 O3' DC3 23 -2.227 -35.914 37.023 1.00 0.00
ATOM 725 HO3' DC3 23 -2.198 -35.362 37.808 1.00 0.00
TER
ATOM 726 HO5' DG5 24 -11.543 -41.964 25.198 1.00 0.00
ATOM 727 O5' DG5 24 -12.183 -41.501 25.743 1.00 0.00
.
.
ATOM 1455 O3' DC3 46 -27.509 22.623 4.002 1.00 0.00
ATOM 1456 HO3' DC3 46 -28.067 23.305 3.622 1.00 0.00
TER
ATOM 1457 N MET 47 -7.175 -28.886 46.154 1.00 0.00
ATOM 1458 H1 MET 47 -7.456 -29.333 47.015 1.00 0.00
.
.
ATOM 3511 O HIE 177 -44.231 7.804 19.703 1.00 0.00
ATOM 3512 OXT HIE 177 -42.739 9.494 19.175 1.00 0.00
TER
ATOM 3513 N MET 178 -34.183 14.573 3.448 1.00 0.00
ATOM 3514 H1 MET 178 -33.328 15.108 3.501 1.00 0.00
.
.
ATOM 5567 O HIE 308 -21.614 -23.924 59.395 1.00 0.00
ATOM 5568 OXT HIE 308 -23.702 -24.376 58.960 1.00 0.00
TER
ATOM 5569 AG AG 309 -50.977 -10.211 13.888 1.00 0.00
TER
ATOM 5570 AG AG 310 -29.668 -11.263 60.213 1.00 0.00
TER
(Na+, Cl- ions and water)
==================================================================
Before creating the prmtop/inpcrd files, I connect the AG atoms to some
residues in the protein and then create the prmtop/inpcrd. There is no
issue in leap while doing this. But when I use ambpdb, the .pdb file is
created wrong.
The reason why I expect this not to affect the calculations is that if I
do no solvate the system with water, ambpdb creates the right .pdb files.
Somehow, ambpdb uses ATOMS_PER_MOLECULE while creating .pdb files if the
system is solvated. Just a note, there is no ATOMS_PER_MOLECULE section if
I do not solvate the system with water.
Also usage of bond in leap is common. That is another reason why I think
this should not affect the calculations. Thanks.
Ilyas Yildirim, Ph.D.
-----------------------------------------------------------
= Department of Chemistry - 2145 Sheridan Road =
= Northwestern University - Evanston, IL 60208 =
= Ryan Hall #4035 (Nano Building) - Ph.: (847)467-4986 =
=
http://www.pas.rochester.edu/~yildirim/ =
-----------------------------------------------------------
On Mon, 2 Jul 2012, Jason Swails wrote:
> On Mon, Jul 2, 2012 at 2:42 PM, Ilyas Yildirim
> <i-yildirim.northwestern.edu>wrote:
>
>> I think this issue was discussed before in amber mailing list but I want to
>> double check on one thing:
>>
>> Assume that you have a protein-DNA system. And assume that you define two
>> heavy
>> atom such as AU or AG which are covalently connected to some of the
>> residues in
>> protein. There is a tutorial in the amber website (created by Ross) that is
>> almost similar to this system except that the heavy atoms are included in a
>> mutated residue, which is created by the user.
>>
>> http://ambermd.org/tutorials/advanced/tutorial1_orig
>>
>> Assume that one does not include these heavy atoms in a mutated residue but
>> explicitly use the 'connect' command in leap to create the covalent bonds.
>> If
>> the 'frcmod' includes all the necessary parameters, leap will create
>> prmtop and
>> inpcrd files without any issues. Savepdb command will also create the rigth
>> sequencing of the system (let's say it is called mol.pdb).
>>
>> Now, if I use 'ambpdb -p prmtop <inpcrd> inpcrd.pdb', the file created is
>> not
>> right (not similar to mol.pdb). This was discussed before (in 2005) and the
>> main issue wass because of the way ATOMS_PER_MOLECULE is defined in the
>> prmtop
>> file. I can understand that because the AG or AU atoms which are connected
>> to
>> proteins explicitly will be included in the proteins and they will not be
>> seen
>> as single ions by ambpdb (like the Na+ or Cl- ions). As a result, TER cards
>> will be put to the wrong places by ambpdb.
>>
>> The question I have is the following: Will this effect the computations or
>> the
>> only effect of this way of creating the prmtop/inpcrd files is on ambpdb
>> putting the TER cards to the wrong places? I will expect that there should
>> not
>> be any issue with the computations but the TER cards placed by ambpdb while
>> creating the .pdb files. Am I correct? Thanks.
>>
>
> I'm not sure exactly what's going wrong here. Is ATOMS_PER_MOLECULE not
> being defined correctly? I know this can happen in tleap sometimes when
> two independent 'molecules' are connected via the 'bond' command, and to
> date I don't think the source of this error has been determined or fixed
> (given there are acceptable workarounds).
>
> If the ATOMS_PER_MOLECULE section is *wrong*, then things can go very wrong
> in both pmemd and sander. A check was introduced in Amber 12 to guard
> against one manifestation of this bug (namely, adding up all numbers in the
> ATOMS_PER_MOLECULE section must sum to NATOM). If this section is
> *seriously* wrong, then pressure scaling may be thrown out of whack (to
> what degree, I'm not sure). pmemd also uses this section to designate PRFs
> (pseudo-residue fragments -- part of the parallel load balancing it does)...
>
> xparmed.py/parmed.py both have a function (setMolecules) that will fix the
> ATOMS_PER_MOLECULE section (or bail out of the prmtop is set up in such a
> way that the ATOMS_PER_MOLECULE section cannot be made correctly). After
> doing this, the TER cards should be put in the correct place.
>
> If the problem lies in ambpdb, on the other hand, then the only issues you
> may see is when you load the PDB back into tleap. Without TER cards
> between molecules, it will recognize polymer units (i.e. nucleic acid
> residues and amino acid residues) as in-chain, and will not appropriately
> pick up the terminal residue (e.g., C- or N-terminus for amino acids).
>
> HTH,
> Jason
>
> --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Candidate
> 352-392-4032
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Mon Jul 02 2012 - 13:00:03 PDT
