[MMTK] Adding/modifying forcefields

[Guido Wagner]

Hi Dave !

You will have to write your own class for the forcefiled. There is no need to 
modify existing code. The example for the harmonic Oscillator ff, that comes 
with MMTK is a good way to get started. After that you should have a closer 
look at the Amber94 implementation. It uses the following approach:

The universe class is derived from a forcefield class which implemets a method 
called evaluatorTerms, which iterates over the objets in the universe and 
builds a list of energyterms to evaluate. An other method (energyEvaluator) 
evaluates the list of energyterms (this functionality is split up in split up 
in subclasses, but I will not discribe this here). An element in the list of 
energyterms consists of a "reference" to the evaluation code (here 
implemented in c) and the necessary data.
In the c part you will find the actual ff code (calculation of energy and 
derivates [ for the energy minimizer ] and lots of "book keeping" code, that 
is important to understand, if you want to be compatible to other MMTK 
modules like the energy minimizer. I think you should try to extend the 
Amber94 ff for your needs. Good Speed !

Guido Wagner
 

Am Freitag, 17. Dezember 2004 13:20 schrieb david.evans at ulsop.ac.uk:
> Hi,
>
> I'd like to use the Amber99 ff, in conjunction with gaff to describe
> ligand binding energetics.
>
> I have, I think, understood how to create new molecules using
> database files in
> .mmtk/Database/Molecules
> and would like to do something similar for forcefields i.e. without
> modifying the base distribution.
>
> Thanks
>
> Dave
>
> _______________________________________________
> mmtk maillist  -  mmtk at starship.python.net
> http://starship.python.net/mailman/listinfo/mmtk