Re: [ccp4bb]: twin refinement with arp/shelxl

[Anastassis Perrakis <perrakis@nki.nl>]

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

> I wish to  refine a merohedric twin to 2.0 A and, as usual, the model
> bias seems very difficult to get
> rid of. I was wondering if someone has already tried to couple arp
> with shelxl, in order to try an
> automatic way of refining the twin ratio and exclude the phase bias
> (the arp_shelxl script available
> in the documentation was prepared for other purposes). Any hint is
> welcome. Thanks

 
I suggest that the easiest approach would be to use DETWIN from ccp4 to
create a detwinned dataset and use that for arp/warp. With 2.0 A
data arp/warp could be able to autobuilt a good model.

You can always go back to shelxl later and continue refinement
in your prefered way, but I think an approximate detwinning will
be OK for refmac/arp/warp.  Maybe try that in a higher resolution
structure first (I think you have a few ...) to see if I am wrong or
not.
We had a coupel of years ago some lack doing that with Anke T.v.S. et al
(Uppsala) but that was with 1.3 data ...

                Tassos

P.S Hope I re-subscribed corrrectly from NKI, but if this ends in the
mail
bin of the list admin, pls fwd to list ...