Hi Miguel,
If you have two molecules each surrounded by additional molecule
copies that make crystal contacts you can align the two crystal lattices
as follows. Align the two original molecules (e.g. using Match Maker
dialog or match command). Then choose all copies of the first molecule
in Model Panel and use the Transform As button. You'll be asked to
choose the molecule to transform these as, so choose the aligned
molecule. That will bring all the crystal contact molecules surrounding
the aligned molecule into correct locations around the aligned molecule.
This method assumes you have saved the crystal copies to a PDB file
and have opened that file. If you simply use the new crystal contacts
option to make the molecule copies then the Transform As will move all
those copies to the exact same position as the original molecule -- not
what you want. This is because the crystal contacts tool is placing the
surrounding molecules by adjusting their transformation matrices rather
than changing their atom coordinates. If you save all copies to a PDB
file and read it back in then the atom coordinates of the copies will be
adjusted and all of the molecules will have the same transformation
matrix and the above alignment procedure will work.
Tom
Miguel Ortiz-Lombardía wrote:
> Hi again!
>
> I had tried this, yes. But the problem is that when I want to compare
> crystal packings the procedure is:
>
> 1. Generate the two ensembles around "central" molecules
> 2. Superpose the "central" molecule (submodel if saved as you say) of
> the first ensemble to the "central" molecule of the second one
>
> Doing it this way only the central submodel of the first ensemble moves,
> not the whole ensemble, as I wish... A single molecule with various
> chains would move as a whole, as I want. I see the 62 chains limitation,
> but normally we don't have such enormous amount of molecules contacting
> an asymmetric unit (of course, with viruses may be different). So, it
> would be nice (to me, I don't know how generally) to have a way to
> transform an ensemble of models into an ensemble of chains.
>
> Thank you for your help!
>
>
> Miguel
>
> 2008/4/30 Tom Goddard <goddard(a)cgl.ucsf.edu <mailto:goddard@cgl.ucsf.edu>>:
>
> Hi Miguel,
>
> You mentioned that you would like to be able to save the crystal
> contact asym units in a single PDB file as separate chains. Why not
> save them as multiple models in a single PDB file? You can do this
> currently with the Chimera menu entry File / Save PDB.... Making
> separate chains requires making new chain names and since only one
> letter names are allowed in PDB format this is limited to about 62
> chains (a-z, A-Z, 0-9).
>
> Tom
>
>
>
>
> --
> http://www.pangea.org/mol/spip.php?rubrique2
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Je suis de la mauvaise herbe,
> Braves gens, braves gens,
> Je pousse en liberté
> Dans les jardins mal fréquentés!
> Georges Brassens
