I looked at BONDS_WITHOUT_HYDROGEN and the indexes pointing to the
BOND_FORCE_CONSTANT entries are different in the two prmtop files, indeed
indicating the different ordering of the force constants. Thus, the
topologies seem to be the same in both files :) I also opened xparmed.py
and printed a few bonds, and they came out the same.
With this two tests coming out fine, it seems things are OK with the two
prmtop files, thank you all for your help!
On Wed, Aug 21, 2013 at 11:29 AM, Jason Swails <jason.swails.gmail.com>wrote:
> On Wed, Aug 21, 2013 at 11:03 AM, Jose Borreguero <borreguero.gmail.com
> >wrote:
>
> >
> > %FLAG
> > BOND_FORCE_CONSTANT
> > %FORMAT(5E16.8)
> >
> > 3.40000000E+02 1.89650000E+02 3.50120000E+02
> >
> > W
> > hile the second once contains:
> > %FLAG
> > BOND_FORCE_CONSTANT
> > %FORMAT(5E16.8)
> >
> > 3.50120000E+02 3.40000000E+02 1.89650000E+02
> >
> > Looks like the values are shuffled, but that would imply that the
> ordering
> > of the atoms has changed. However, the two prmtop files are indentical
> when
> > looking at the %FLAG ATOM_NAME section.
> >
>
> The shuffling does not imply that the atom ordering has changed, although
> it is admittedly misleading. Determining the equivalence of two topology
> files is actually pretty difficult. All this shuffling means is that the
> the bond force constant are shuffled (not much information there ;)).
> These values are addressed via pointers in BONDS_WITHOUT_HYDROGEN and
> BONDS_INC_HYDROGEN (see http://ambermd.org/formats.html for more
> information). The BONDS_WITHOUT_HYDROGEN and BONDS_INC_HYDROGEN assign an
> index into the BOND_FORCE_CONSTANT and BOND_EQUIL_VALUE arrays for each
> bond (that is defined by 2 atom indexes). Therefore, if you shuffle the
> BOND_FORCE_CONSTANT section, all you have to do is adjust the assigned
> indices in BONDS_WITHOUT_HYDROGEN and BONDS_INC_HYDROGEN.
>
> You can use the "print*" commands in ParmEd (printBonds, printAngles,
> printDihedrals, printDetails) to verify this (or even more definitively,
> use sander to compute the energy using both topology files and see that
> they are identical).
>
> 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
Received on Wed Aug 21 2013 - 11:00:03 PDT
