Re: [ccp4bb]: NCS with molecular replacement

[Edward Berry <eaberry@lbl.gov>]

***  For details on how to be removed from this list visit the  ***
***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***



Jeff Taylor wrote:

> I have a rather sticky molecular replacement problem and I'm looking for
> helpful ideas.  My protein crystallizes in the C2 space group, although
> homologs of this protein crystallize in P61.  I mention the P61 crystals
> because when I try a self rotation on the C2 data I get peaks on the
> Kappa=60, 120, and 180 sections which look just like those from the P61
> data.  The cell dimensions for the C2 crystal are large and from the
> Matthews number, I'm expecting at least 6 molecules in the asymmetric
> unit (whereas the P61 data has one molecule/asymm), and so the signal in
> a cross rotation search will be small.  Has anyone worked out a way to
> take advantage of a NCS in a molecular replacement search?
> 

You could use a multimer as the search model. This would make sense
if it is a multimer in solution so that the same multimeric arrangement
is present in both crystals.

That would be for molecular replacement with coordinates for search
model or cut-out density from a phased crystal. For doing cross rotation
function between two unphased crystals, I think you automatically get the
advantage of any common multimeric structure: say one crystal contains
three dimers in the asu and the other has the same dimer but with the
2-fold on a crystallographic axis. Patterson cross peaks between the 
two monomers, which could be seen as self peaks within the dimer, will
superimpose when the orientation is correct. You might want to use
a larger sphere radius, i.e. based on size of the dimer not monomer.

I think with the rotation function it is the number of copies in the unit cell
rather than the number in the asymmetric unit that determines your signal.
So if there is no common multimer the correlation signal would come from 
superimposing 1/6 of the P61 crystal on 1/12 of the C2 crystal, whereas if 
there is a common dimer you would be superimposing 1/3 on 1/6 for maybe
4 x better signal to noise.
-- 
Edward A. Berry, MailStop 3-250
Lawrence Berkeley National Laboratory
1 Cyclotron Road, Berkeley, CA 94720
Phone: +1-510-486-4335
Fax: +1-510-486-6059
Jfax +1-530-323-9836 (you send fax, I receive email)
e-mail: EABerry@lbl.gov