On Fri, Oct 30, 2015 at 11:02 AM, Alan <alanwilter.gmail.com> wrote:
> I can tell you the only component that did not differ: BOND.
>
​A fair number of the terms seem to match: proper torsions, improper
torsions, L-J (both -1.3323 kcal/mol), and electrostatics (both are -269.7
kcal/mol). The 1-4s clearly don't match, nor do the angles [with
urey-bradleys] and CMAPs. It seems to me that the rest do. I will try to
look into why these conversions are failing for a bunch of the potential
terms...
Sander:
> BOND = 1.3351
> UB = 0.3669
> 1-4 NB = 2.1041
> EELEC = -269.7122
> EKtot = 0.0000
> ANGLE = 14.1158
> IMP = 0.3344
> 1-4 EEL = 278.1614
> EHBOND = 0.0000
> EPtot = 39.1266
> DIHED = 14.2773
> CMAP = -0.5239
> VDWAALS = -1.3323
> Etot = 39.1266
>
> GMX kJ/mol kcal/mol)
> Bond 5.58589 1.335059751
> U-B 60.5954 14.48264818
> Proper Dih. 59.7363 14.27731836
> Improper Dih 1.3993 0.334440727
> CMAP Dih. -17.3579 -4.148637667
> LJ-14 94.6652 22.62552581
> Coulomb-14 1163.8 278.1548757
> LJ (SR) -5.57431 -1.332292065
> Coulomb (SR) -1128.45 -269.7060229
> Potential 234.4 56.02294455
> Kinetic En. 0.168643 0.040306644
> Total Energy 234.569 56.06333652
>
> On 30 October 2015 at 01:40, Jason Swails <jason.swails.gmail.com> wrote:
>
> > (I actually have a test in ParmEd that goes GROMACS -> AMBER
> > using the CHARMM FF and get the same results in both programs).
> >
>
> Can you tell more about this test? Where to find?
>
​You can find the format conversion tests in the test suite:
https://github.com/ParmEd/ParmEd/blob/master/test/test_format_conversions.py#L56-L192
All of the energy comparisons are done using OpenMM, but I've validated all
of the OpenMM energies against the format's parent program (and OpenMM
gives the same results as GROMACS when starting from a GROMACS topology).
All the best,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Oct 30 2015 - 09:00:04 PDT
