[phenixbb] AutoMR:Multi-subunit model

[Nathaniel Echols]

On Mon, Jun 21, 2010 at 12:49 PM, S. Saif Hasan <sshasan at purdue.edu> wrote:

> I have a membrane protein complex made up of 8 different polypeptides
> present in
> two copies each (16mer). The asymmetric unit is the octamer, as determined
> from
> previous crystallographic studies. To carry out molecular replacement with
> the
> phases from the published structure in AutoMR, should I specify the
> coordinates
> of the search model and its overall sequence identity as one file or should
> I
> break it down for each subunit ?


If it doesn't have any large conformational changes, using the entire
octamer might work.  It will certainly be much faster than placing each
chain separately, so it's worth trying first anyway.  You should run
rigid-body refinement afterwards with a rigid group for each chain.  If you
can't place the octamer, try searching for smaller sub-assemblies before you
break it down into individual chains.

Also, for the asymmetric unit contents, should
> I specify the sequence or mass of each subunit separately (as component 1,
> 2 and
> so on) keeping the number of copies fixed at 1 ? There is no NCS.
>

I think you can just specify the overall mass - regardless of whether you're
searching for the entire complex or one chain at a time.

-Nat
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20100621/3c4e48d3/attachment-0003.htm>