[phenixbb] questions about ncs copy in autobuild

Vellieux Frederic Frederic.Vellieux at ibs.fr
Thu Jul 22 00:28:54 PDT 2010


vandana wrote:
> i have one 3.27 A data of one of my protein. it is a multidomain 
> protein . i  used molecular repacement (phaser ) to find individual 
> domains i got density for 3 domains.
> with last 4th domain i am getting solution with 5 clashes and density 
> is nort clear for that . so i used phenix find helix and strands 
> module i got 43 fragments model . now i dont know how to proceed 
> further how i can start fitting. suggest me for further step .
> ------------------------------------------------------------------------
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb
>   
Hi,

What happens when you carry the molecular replacement search using the 
entire protein (4 domains) as the search model? This is only applicable, 
of course, if the 4 domains have the same relative orientation in the 
model and in your structure.

Sometimes, one has to resort to real space molecular replacement in 
order to position domains, an example is given in Cosenza et al. (2002), 
J. Mol. Biol. 318, 1417-1432 (one domain had to be placed "in real 
space" after suitable density modification, i.e. "protein flattening" - 
don't know if it is possible to do that in Phenix at the moment).

Further, you may have one domain that is disordered (or much less 
ordered) than the other 3. An example is in Xia et al. (1987), PNAS 84, 
2629.

Fred.



More information about the phenixbb mailing list