Re: [AMBER] Topology issue with MMPBSA.py

From: Daniel Roe <daniel.r.roe.gmail.com>
Date: Thu, 19 Jun 2014 15:24:40 -0600

FYI you can use either cpptraj or parmed.py to check the number of bonds to
an atom in your topology, e.g.

cpptraj:
> bonds .12
# Mask [.12] corresponds to 1 atoms.
# Bond Kb Req atom names (numbers)

     107: 570.00 1.229 :1.C :1.O (12,13)
     108: 490.00 1.335 :1.C :2.N (12,14)
     111: 317.00 1.522 :1.CA :1.C (5,12)

parmed.py:
> printBonds @12
Atom 1 Atom 2 R eq Frc Cnst
     12 C ( C) 13 O ( O) 1.2290 570.0000
     12 C ( C) 14 N ( N) 1.3350 490.0000
      5 CA ( CT) 12 C ( C) 1.5220 317.0000

-Dan


On Mon, Jun 16, 2014 at 7:37 AM, JPB <jpb.q4md-forcefieldtools.org> wrote:

> Hi Jason,
>
> Thanks for quick answer and insight.
>
> To clarify the warning message itself:
> C: 11 1;
> means the C atom number 11 has only 1 bond, doesn't-it?
> I was not clear where the 1 was coming from.
>
> I had already checked my topology with xleap and vmd
> and this is one of the reasons this message was so disturbing.
>
> Moreover, all the other carbon atoms in the structure are described
> by CA type as well, each of them have 3 bonds, and none of them causes
> problem.
> I also tried a similar computation with the same procedure on
> ethylbenzene and everything went smoothly, no warning message.
> It really looks like an Sp2 (in vinylbenzene) carbon at this
> particular position
> triggers the output of the warning message and not a Sp3 one (in
> ethylbenzene).
> As I said, it is still kind of disturbing.
>
> Cheers.
> JPB.
>
> Jason Swails <jason.swails.gmail.com> a écrit :
>
> > On Mon, 2014-06-16 at 12:29 +0200, JPB wrote:
> >> Dear All,
> >>
> >> I am getting use to running MMPBSA.py jobs.
> >> For this purpose I am working on vinylbenzene.
> >> In the _MMPBSA_complex_gb.mdout files the following error message
> appears:
> >>
> >> bad number of bonds to C: 11 1; using default carbon parameters
> >>
> >> I did not manage to solve this problem.
> >> Another thing which is a little confusing is the numbering of atoms in
> this
> >> error message.
> >> In the _MMPBSA_complex.pdb file, atom number 11 is hydrogen, not carbon.
> >> After several attempts on different systems and with different atom
> >> orderings, it seems that the atom numbering starts from 0 in the error
> >> message and not from 1 like in the pdb file.
> >> Is this correct? Can anybody confirm it?
> >
> > Yes, this is correct. This message is printed by NAB (a C-based
> > language), in which arrays are indexed from 0 instead of 1 (as in
> > Fortran-based programs like sander and pmemd).
> >
> >>
> >> Here is the procedure followed for this computation:
> >> 1. generation of a topology for solvated vinylbenzene with tleap.
> >> 2. generation of a topology for vinyl with tleap (dry topology).
> >> 3. running a short molecular dynamics of vinylbenzene with pmemd in
> >> TIP3P water
> >> using the solvated topology.
> >> 4. rewriting the trajectory removing water molecules with cpptraj.
> >> 5. running MMPBSA.py on the trajectory without waters using the dry
> >> topology.
> >>
> >> It seems the carbon atom making trouble is the one at the end of the
> >> vinyl group to which two hydrogen atoms are linked.
> >> (cross-checked on other compounds)
> >> Any help to solve this issue would be deeply appreciated.
> >
> > MMPBSA.py can actually automate step 4 for you, in case you cared.
> >
> > In any case, this seems like a warning, not an error, but it is a
> > concerning warning. Atom number 12 here is recognized as a carbon atom,
> > but only 1 bond to that atom is detected. Carbon forms at least 2 bonds
> > (think alkyne or allene) in any biologically relevant system. So if one
> > of your carbon atoms has only a single bond you need to find out why and
> > fix it.
> >
> > However, this error is printed out by the LCPO calculation in
> > mmpbsa_py_energy, which is actually ignored entirely by MMPBSA.py in
> > AmberTools 12 or 14. The SASA calculation is dispatched to cpptraj
> > (which determines the element based on the atomic number rather than
> > based on the atom type name). So if you are confident that there is no
> > such error in your system, you can safely ignore this warning.
> >
> > As a hint, visualizing your system (using your prmtop in VMD, _not_ by
> > converting to a PDB first) can help you detect missing bonds or other
> > oddities in your topology.
> >
> > HTH,
> > Jason
> >
> > --
> > Jason M. Swails
> > BioMaPS,
> > Rutgers University
> > Postdoctoral Researcher
> >
> >
> > _______________________________________________
> > 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
>



-- 
-------------------------
Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 201
Salt Lake City, UT 84112-5820
http://home.chpc.utah.edu/~cheatham/
(801) 587-9652
(801) 585-6208 (Fax)
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jun 19 2014 - 14:30:04 PDT
Custom Search