[gmx-users] Entropy calculation from normal modes

[Marc Baaden]

Hi David,

you are right, I am currently doing an MM_PBSA analysis from my data.
We both have Gromacs and Amber, and I started from the Amber8 MM_PBSA
implementation, which I adapted to use Gromacs MM energies, and run PB/SA
from my Gromacs trajectories. SA is fine, but PB currently has (obviously)
not the right parameter set, hence my question about radii and charges.
Ideally I'd like a charge and radius file that tells Delphi to assign
PARSE parameters to a PDB file with ffgmx atom types.

I could have gone the route to convert my gmx trajectory to Amber, but I
don't think this is consistent, in particular concerning the MM energy
part. That is unless you are able to create an Amber topology with gmx
forcefield parameters !?

I think this reveals a general difficulty with these approaches. One
usually has a lot of different ways to calculate the different energy
terms, all leading to a different end result. To choose the "right"
way, one would have to run a number of testcases where the (experimental)
result is well known and characterised. And even then I am not sure that
one would reach a unique answer.

Marc

>>> david.evans at ulsop.ac.uk said:
 >> It sounds as if you are trying to recreate the MM_PBSA approach from
 >> Amber. That package has a nice perl script which does all of these
 >> things (PB calculations, normal more analysis) automatically. The
 >> approach I'm using at the moment is to run my trajectories in
 >> Gromacs, convert them to Amber format and analyse with Amber. Of
 >> course this requires using a ff which runs in Amber, (and paying for
 >> a license for the non-GPL Amber package), which may cause you more
 >> problems than you solve.=20
-- 
 Dr. Marc Baaden  - Institut de Biologie Physico-Chimique, Paris
 mailto:baaden at smplinux.de      -      http://www.marc-baaden.de
 FAX: +49 697912 39550  -  Tel: +33 15841 5176 ou +33 609 843217