On Sun, Oct 4, 2015 at 2:32 AM, conor parks <coparks2012.gmail.com> wrote:
> Thanks for the help. Just to clarify, amber in no way changes the SPC/e or
> tip3p paramaters during the overall force field parameters determination,
> correct? I am just trying to rationalize in my head how it is possible to
> simulate using OPLS/TIP3P and amber/TIP3p if the mixing rule are different
> but the water force field parameters are the same.
One thing I'll point out is that for both SPCE and the TIPnP water models,
the hydrogen Lennard-Jones are zero. What this means is that the only
Lennard-Jones terms that exist in pure water are between oxygen atoms. One
of the nice results of this is that combining rules *do not matter*.
Geometric and arithmetic means are identical for an atom type interacting
with itself (sqrt(sig*sig) == (sig+sig)/2). So TIP3P and SPCE are the same
in OPLS as they are in force fields that use the Lorentz-Berthelot
combining rules (protein-water interactions differ, but that is to be
expected as the force fields themselves are different, but the water-water
interactions are still the same).
> Perhaps the force field
> themselves were determined via parameter optimization in conjunction with
> the tip3p model and SPC/E model. Can anyone clarify if this is the case for
> amber?
The definitive source for this information is in the force field
parametrization papers. But the charges are derived via RESP, so the
solvent does not come in at all there (unlike CHARMM, which I believe
optimizes charges for use in their modified TIP3P model). I don't *think*
the parameters are heavily optimized inside explicit solvent (i.e., to a
specific solvent model), but the articles are the place to find this
information for sure.
I am looking in the GAFF file, but I don't see any reference to
> SPC/e water atoms. However, in the mod4 section the epsilon value for
> oxygen is 0.1520, but I believe the original 1983 paper had it as 0.1521.
> Just trying to make sure I am not missing something.
>
The water models are in frcmod.<water model> -- so look for frcmod.spce
(and the default TIP3P parameters are in parm10.dat -- look for OW and
HW). I can't speak to the discrepancy between 0.1520 and 0.1521, but I can
verify that TIP3P has epsilon=0.1520 for the oxygen in the Amber force
fields.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Oct 04 2015 - 08:00:03 PDT
