[MMTK] Amber parameters for ligands

[Konrad Hinsen]

On Nov 19, 2008, at 17:46, Lars Skjærven wrote:

> so the following would be the way to go?
>
> for each atom in ligand:
>      atom.amber_charge = 'charge_from_amber_prep_file'
>      atom.amber_atom_type = 'atom type from amber prep file'
>
> Would MMTK then know the topology, and thus we would be "set to go"
> for minimization?

No, you need to tell it where the bonds are. However, you can create  
the bond structure through a MoleculeFactory, that's why I wrote that  
a database file is not strictly necessary. It may well be the easiest  
way for you, of course.

> A more general way would be cooler though. What would I need in order
> to build a script that automatically builds MMTK database files from
> amber topology files ? Will it not be rather straight forward? I guess
> there is a reason for why you did not do it fully automatically, so
> there is something I don't see yet.

For a single molecule, the conversion would be rather straightforward  
and could very probably be done automatically by a script. What I  
converted many years ago was the complete database of amino acid and  
nucleic acid residues. I had to do the mapping from Amber residue  
names to MMTK group names manually, which is why the conversion was  
not fully automatic.

Konrad.
--
---------------------------------------------------------------------
Konrad Hinsen
Centre de Biophysique Moléculaire, CNRS Orléans
Synchrotron Soleil - Division Expériences
Saint Aubin - BP 48
91192 Gif sur Yvette Cedex, France
Tel. +33-1 69 35 97 15
E-Mail: hinsen at cnrs-orleans.fr
Web: http://dirac.cnrs-orleans.fr/~hinsen/
---------------------------------------------------------------------