I hadn't thought much about how the normal mode entropy would be
decomposed, but decomposition support was added to the nmode program in
2009. I believe mm_pbsa.pl supports normal mode entropy decomposition, but
I have never used it (and never even looked into it).
Look at the Amber 12 manual for more instructions (mm_pbsa.pl). The code
is pretty straightforward ($AMBERHOME/src/nmode/decomp.F90), so you should
be able to figure out how the contribution for each normal mode (vector) is
attributed to each residue.
HTH,
Jason
On Sat, Feb 16, 2013 at 1:06 PM, Aron Broom <broomsday.gmail.com> wrote:
> Hi AMBER users,
>
> I'm wondering if it is theoretically possible to decompose the entropy term
> from an MM PBSA/GBSA calculation to determine the contribution of each
> residue?
>
> I realize the script currently does not support this, and the manual makes
> that quite clear. I also believe I understand that in taking in the whole
> complex there would be no way to actually do this decomposition easily.
> What I'm wondering is, from a theoretical perspective, if one did the
> normal alignment of the whole complex to remove those irrelevant rotations
> and translations, and then stripped out all but the residue of interest,
> and performed the entropy calculations (Quasi-harmonic or nmode) on just
> that reduced trajectory (which has been pre-aligned as part of the whole
> complex), would you get something sensible? That is, would doing this for
> each residue give a final result that more or less added up to what you
> would get from doing the whole complex (I realize there is a lot of
> uncertainty in entropy calculations as it is, and so there would be a
> tremendous propagation of error involved)?
>
> I'm asking about this because MM/PBSA calculations seem to offer a very
> nice advantage over other methods for determining say ligand binding energy
> (umbrella sampling, thermodynamic integration etc..) in that one can get
> these per residue decompositions of the energy terms, and thereby get a
> clue as to how the protein for instance, might be re-engineered for better
> binding, whereas methods like umbrella sampling would essentially require a
> guess, followed by another whole analysis of the mutated complex in order
> to test only one potential mutation.
>
> Thanks,
>
> ~Aron
>
> --
> 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
>
--
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 Sat Feb 16 2013 - 12:30:02 PST
