[MMTK] AMBER force field

Konrad Hinsen hinsen@cnrs-orleans.fr
26 Sep 2001 18:06:47 +0200


Berit Hinnemann <Berit.Hinnemann@fysik.dtu.dk> writes:

> I did this by taking all parameters of the 'amber_parm94' file and
> simply replacing names, i.e. C with WX, CT with WT and so forth. 
> 
> I compared the different energies by using 'universe.energyTerms()' and
> all energies agree, except for the dihedral cosine sum. I also tried to

This is well possible. If you look the Amber force field definition in
detail, you will find that the improper dihedral terms depend on the
alphabetical ordering of atom names. For every improper dihedral
configuration (three atoms all bound to a central forth one) there are
three different dihedral angles, of which only one is taken, and the
selection is based on alphabetical order. The difference should in
general be small though.

> Another question: When I create the universe and specify the force field
>      universe=InfiniteUniverse(Amber94ForceField())
> 
> is it possible to specify a modification file for AMBER? From how I
> understood the code it should be something like
> Amber94ForceField(modfile='amber_modfile'), is that right? Can I specify

The syntax is:

   Amber94ForceField(mod_files=['amber_modfile'])

as you may need several modification files.

> the path and is the default path the one where 'amber_parm94' and
> 'amber_parm94.heme' are located?

The default path is the current directory, but of course you can specify
full paths.

> Another thing: Is it possible to simulate a chain, e.g. consisting of
> glycine residues in a periodic universe such, that the chain is repeated
> in x-direction? I would create the chain with the options
> 'PeptideChain(n_terminus=0, c_terminus=0, circular=1)'. Is the bond
> between the last residue in one unit cell and the first residue of its
> repetition formed correctly, i.e. as a bond with the smallest possible
> length?

I admit that I have never tried this, but it should work, as all distances
are calculated as minimum-image distances.

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen@cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------