AMBER Archive (2005)

Subject: Re: AMBER: MM_PB/GBSA

From: Thomas E. Cheatham, III (cheatham_at_chpc.utah.edu)
Date: Thu Sep 29 2005 - 23:04:46 CDT


> I am dealing with some protein-inhibitor systems and want to calculate
> the binding energy between the receptor and the drugs. However, the
> difference between the IC50 values from experiments is very very small,
> for instance, 0.16um for one, and 0.1um for another. So, I was wondering

  delta_G = -RT ln K

You can plug in the differences in the equilibrium constant to estimate
approximately what the change in relative free energies is using the
relationship between free energy and equilibrium constant shown above.
As you mention, the difference is very very small. For each order of
magnitude in equilibrium constant, this is worth a mere ~1.4 kcal/mol at
300K. A difference of .1 to .16 in Keq (or loosely IC50) is less than a
factor of 2, so only on the order of ~0.3 kcal/mol. This is less than
thermal fluctuation (~0.6 kcal/mol at 300K) and a very small fraction of
the fluctuation in potential energy represented in our simulations.
Assuming complete Boltzmann sampling, in principle it might be possible to
observe such small differences in our calculated estimates. In practice,
given sampling errors, differential substate sampling of the protein and
ligand, and inherent errors in the representation (ranging from estimates
of the entropy to fluctuations in the energy/potential to neglect of
specific hydration in the MM-PBSA estimates), it is unrealistic to expect
kcal/mol accuracy let alone < 1.0 kcal/mol accuracy.

As I think about it, the key is not if we can reproduce such differences
with simulation (as we likely cannot), but can the simulation of the
complexes provide insight into differential interactions, hydration, and
so on and perhaps guide you towards design/prediction of better
inhibitors. This limitation in accuracy is not limited to simulation
alone (although it is significant in simulation and likely greater than
+/- 1.0 kcal/mol); if you consider experimental error, differences in Ki
vs. IC50 vs. binding constants, let alone who does the experiment, likely
you will also see errors on the order of a kcal/mol or ~+/- an order of
magnitude binding.

[Feel free to correct me if I am wrong in my interpretations!]

\-/ Thomas E. Cheatham, III (Assistant Professor) College of Pharmacy
-/- Departments of Med. Chem. and of Pharmaceutics and Pharm. Chem.
/-\ Adjunct Asst Prof of Bioeng.; Center for High Performance Computing
\-/ University of Utah, 30 S 2000 E, SH 201, Salt Lake City, UT 84112
-/-
/-\ tec3_at_utah.edu (801) 587-9652; FAX: (801) 585-9119
\-/ BPRP295A http://www.chpc.utah.edu/~cheatham

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu