Hi. Basically, you should make sure that the reference states with respect
to which you are measuring free energies are the same in the experimental
and computational studies. The usual experimental reference is 1 molar,
which is equivalent to a volume of 1661 angstroms. In that case, if you are
doing TI with the so called double-decoupling method, DDM, i.e, using
restraints so that the ligand remains in the binding site throughout the
integration, then you need to correct your binding free energy by
-KT ln (V_bs / 1661),
where V_bs is the volume of the binding site.
Or, if you are not using restraints and the ligand is free to move all over
space (the double-annihilation method, DAM), then the correction is
-KT ln( V_p / 1661),
where V_p is the volume of the simulation box in the case where you have
the complex (ligand+protein).
In any of these cases you should make sure that your integrations are well
converged.
Here are some papers that cover the foundations of this standard state
issue:
1) Gilson, M. K.; Given, J. A.; Bush, B. L.; McCammon, J. A. Biophys. J.
1997, 72, 1047–1069.
2) Boresch, S.; Tettinger, F.; Leitgeb, M.; Karplus, M. J. Phys. Chem. B
2003, 107, 9535–9551.
3) General, I. J. JCTC 2010, 6, 2520-2524.
Ignacio
On Wed, Nov 28, 2012 at 1:05 PM, manikanthan bhavaraju <
manikanthanbhavaraju.gmail.com> wrote:
> Dear Amber Developers,
>
> I have performed TI calculations on protein+ligand system using ff99
> and amber8. Now I am interested to compare my results with the
> experimental values. After some literature review I have found that
> people have experimentally determined dissociation constants. I have
> few doubts about comparing TI results with the experimental values.
>
> I have calculated the experimental free energy for protein-ligand --->
> protein + ligand (3.9 kcal/mol). The TI's deltaA value is 9.2
> kcal/mol
>
> Note: TI's binding deltaA was -9.2 kcal/mol
>
> I am assuming that the experiment deltaG as deltaG(std). In order to
> compare TI value with the experimental value I have to calculate
> deltaG(std).
>
> deltaG (std) = deltaG - RTlnK
>
> What information I have to extract from the TI simulations to
> calculate K value? or What terms should be considered ?
>
> If my assumption is wrong please correct me. Is there any article
> that I can refer to ?
>
>
> Thanks,
> mani
>
> _______________________________________________
> 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 Nov 28 2012 - 11:00:02 PST
