khinsen at cea.fr khinsen at cea.fr
Mon Jul 18 17:21:25 CEST 2005

On Jul 18, 2005, at 16:50, Matias Saavedra wrote:

> personal powerbook G4). So i switched to subspaces.
> With RigidMotionSubspace my computer take 3 hours to
> calculate the Normal Modes of my 300 aa protein. Is

What are the rigid units that define your subspace?

> there a way to reduce the calculation time, but
> keeping all the detailed analysis given by
> Amber94ForceField ?

The subspace you choose is as important as the force field in deciding 
the amount of detail you can expect in your calculations.

> Is the RigidMotionSubspace detailed enough to consider
> all the interactions forces in the protein for the
> calculations?

That depends very much on what your rigid units are, plus on the 
properties of your protein and on the aspects you wish to study.

Note that for studying active sites, it sounds reasonable to have 
smaller rigid units (if any) around the active sites and bigger ones 
further away.

> In the other hand, it is posible to realize a
> minimization in MMTK and obtain the newly generated
> structure conformation ? (this is to perform just one
> minimization in a structure, and be able to save the
> minimized structure in order to use it later without
> the need of minimizing again).

Yes, certainly. Just save the whole universe:

	save(universe, 'my_universe')

and then later load it back:

	universe = load('my_universe')

instead of constructing it from a PDB file.

Don't try to save the minimized configuration in a PDB file for later 
reloading: the precision of the positions in a PDB file is not enough 
to describe the minimum well enough that no further minimization is 

Konrad Hinsen
Laboratoire Léon Brillouin, CEA Saclay,
91191 Gif-sur-Yvette Cedex, France
Tel.: +33-1 69 08 79 25
Fax: +33-1 69 08 82 61
E-Mail: khinsen at cea.fr

More information about the mmtk mailing list