Re: [AMBER] MMGBSA/MMPBSA Decomposition Analysis

From: Jason Swails <jason.swails.gmail.com>
Date: Wed, 27 Feb 2013 21:02:36 -0500

On Wed, Feb 27, 2013 at 7:44 PM, Aron Broom <broomsday.gmail.com> wrote:

> Are the simulations for MM/PBSA done in explicit solvent, and those for
> MM/GBSA done in implicit solvent? This should be the case, that is, you
> shouldn't use the same simulation to do both.


I don't really agree with this. It's common practice to do MM/GBSA
analysis on simulations done in explicit solvent, and there are a number of
papers out there that do exactly this. In fact, some directly compare
MM/GBSA with MM/PBSA results performed on the same ensemble (generated from
explicit solvent dynamics).

Just like PB is an approximation for calculating the mean solvation free
energy of a fixed structure (so you don't have to average out all solvent
DOF), GB is an approximation to PB (and therefore an approximation to the
same quantity).

 Also keep in mind that for
> your MM/?BSA input file, the surface tension term will potentially be
> different. For MM/PBSA the default is something like 0.072, whereas if you
> do GBSA simulations, and turn on gbsa (gbsa=1 in the config file) the
> default is 0.005. Moreover, if you didn't explicitly turn on gbsa, it is
> off by default, and so the actual surface tension term you should use for
> your MM/GBSA calculations is 0.
>

These defaults are correctly stated for sander. For MMPBSA.py, though,
gbsa is on by default and the surface tension is 0.0072 kcal/mol/Ang^2.
 The default PB non-polar solvation method in AmberTools 12 is quite
different -- it uses the approach that decomposes the non-polar part into a
repulsive hydrophobic and an attractive dispersion term (the bug I was
referring to earlier actually neglected this attractive term in certain
instances).

The full list of (correct) MMPBSA.py defaults are listed in the AmberTools
12 manual with errata listed on http://ambermd.org/doc12/ (some of the
updated PB defaults were accidentally omitted from the published version).

HTH,
Jason

-- 
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 27 2013 - 18:30:02 PST
Custom Search