I think you should perhaps do some literature searching concerning the
expected error for MM/PBSA. I'm not sure how much your binding sites
differ, but I think that with very good statistics and a forcefield that
describes your system well, you are looking at an error of +/- 1 kcal/mol
as a good result. Based on your description it doesn't sound like you are
making an extreme effort for highly accurate results, so I'd expect your
error to be > 1kcal/mol, perhaps by quite a lot.
~Aron
On Thu, Mar 8, 2012 at 2:10 PM, mish <smncbr.gmail.com> wrote:
> Well, I do not expect quantitative results from these calculations. I am
> doing these calculations to distinguish between multiple binding sites (for
> same ligand) in a system. Experimentally there not any evidence of
> difference in binding energies for those sites but from Mm/PBSA I had found
> two groups of binding site with diff binding energies. Now when I set salt
> concentration there binding strength is reverse (weak binding site becomes
> strong and vice-versa). This is where I got doubtful about the results
> because including salt changes the qualitative meaning of results.
>
> /Mish
>
>
> On Thu, Mar 8, 2012 at 7:00 PM, Thomas Cheatham III <tec3.utah.edu> wrote:
>
> >
> > > I have question related to salt the salt concentration and ionic
> strength
> > > values used in MM/GBSA and MM/PBSA calculations, respectively.
> Initially
> > I
> > > was using perl script and I did not care much about these two values (
> so
> > > they were set to zero). Now when I switched to new python script, I
> > noticed
> > > in the manual that value of saltconc was set to 0.15 (for IGB=5) and
> 0.10
> > > (for igb=2) in sample input file. Also istrng=0.15 is mentioned for PB
> > > calculations.
> > >
> > > I want to know, if this should effect the binding energies too much ?
> and
> > > also the reason to have those values (when default is 0.0) there in
> > sample
> > > input files of manual ? I looked for experimental detail in my case and
> > > what I can see is ITC measurement were done in 100mM Tris buffer at pH
> > 7.5
> > > . Can you please suggest me optimal (theoretically) values of saltconc
> > and
> > > istring to compare these experimental values taken in 100mM Tris buffer
> > at
> > > pH 7.5 ?
> >
> > These are rather quick calculations, so why not check yourself? My
> > expectation is that these changes will not have a drastic effect on the
> > calculated values. Also, it makes sense that the sample inputs
> > (which represent actual sample inputs people actually ran with) include
> > salt since we do not live in a no-salt environment.
> >
> > Often MM-PBSA is treated as a black box when in fact the results will
> > depend on choices made so it is a good idea to experiment and understand
> > the implications of your particular choices, particularly if the results
> > seem unreasonable. The black-box nature of MM-PBSA can hide errors.
> >
> > As normal binding constants are in the micromolar to nanomolar range,
> > based on dG = -RT ln K you expect binding free energies in the -8 to -12
> > kcal/mol range. It's more complicated due to rotational/translational
> > entropy losses upon binding, loss of vibrational freedom, etc and other
> > factors (which may not be included, or included correctly in the MM-PBSA
> > approach). However, if you calculate a binding constant of -100
> kcal/mol,
> > something is wrong... See papers and reviews by Gilson on the gory
> > details of dG binding.
> >
> > --tec3
> >
> > _______________________________________________
> > 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
>
--
Aron Broom M.Sc
PhD Student
Department of Chemistry
University of Waterloo
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Mar 08 2012 - 11:30:04 PST
