Hi Almudena,
The possibility of sub-optimal data quality has not been mentioned, apart from radiation damage by Goerge Sheldrick.
If you have high multiplicity and think radiation damage could have been the problem, you could try to remove the later images and see if that helps.
Did the images look “good”? Were there ice-rings or other evidence of sub-obtimal flash-freezing? If you have some more crystals, it might be worth taking a few room temperature images and if these look better, try a different cryo-protection strategy.
Or perhaps there was a small satellite, or the spot-shapes were “ugly”?
Modern data processing can get data with acceptable statistics out of what I would have called not-so-good images, but I find refinement R-factors are ofther higher afterwards.
So perhaps trying more crystals or a collection strategy that avoids radiation damage might improve your results.
Of course, how much effort you want to put in depends on the interest of the mutant...
Like Eleanor says, why use molecular replacement if the spacegroup is the same and the cell parameters are very similar? - it’s overkill; I would just have done a few cycles of rigid body refinement. Of course MR doesn’t hurt, although you risk the program putting your new coordinates in a different place in the unit cell, while it may be more convenient having them in the same place as the native.
Mark J van Raaij
Dpto de Estructura de Macromoleculas
Centro Nacional de Biotecnologia - CSIC
c/Darwin 3
E-28049 Madrid, Spain
tel. (+34) 91 585 4616
http://wwwuser.cnb.csic.es/~mjvanraaij
> On 31 Jan 2016, at 13:14, Almudena Ponce Salvatierra <[log in to unmask]> wrote:
>
> So, the "native" , say, space group is P43 21 2. While collecting the data I processed in P43 21 2 and in P41 21 2. Phaser gave the solution in P43 21 2, despite I asked it to search for all the space groups within that point group. Twinning... I would hope there's not... I could check but since the processing worked so well I was not expecting such a thing.
>
> Depending on the refinement strategy, I get mostly R free around 0.36 and R work aroun 0.29-0.30. But I can also get 0.39 and 0.30, this looks to me more like a "big gap".
>
> The electron density looks fine to me. I mean, it is mostly the same structure except for those mutations.
>
> 2016-01-31 12:19 GMT+01:00 Keller, Jacob <[log in to unmask]>:
> Wrong space group? Twinning?
>
>
>
> But actually, 30 is not that high—and what is the R gap?
>
>
>
> JPK
>
>
>
> From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of Almudena Ponce Salvatierra
> Sent: Sunday, January 31, 2016 4:22 AM
> To: [log in to unmask]
> Subject: [ccp4bb] refinement after MR
>
>
>
> Dear all,
>
>
>
> I was happy to collect some data at 2.5 resolution in BESSY on Thursday, of a complex that I should be able to solve by MR (the model structure is available and what I'm working on carries two or three mutations).
>
>
>
> I run Phaser, it gave a solution with TFZ of around 16 and LLG close to 200. However when I start the refinement the R factors stay quite high... what I mean is that they don't drop below 30 and on the other hand there's a gap between the Rfree and the Rwork that is large I would say..
>
>
>
> Any idea of what may be going wrong?
>
>
>
> Thanks a lot,
>
>
>
> best wishes,
>
>
>
> Almu
>
>
>
> --
>
> Almudena Ponce-Salvatierra
>
> Macromolecular crystallography and Nucleic acid chemistry
>
> Max Planck Institute for Biophysical Chemistry
>
> Am Fassberg 11 37077 Göttingen
>
> Germany
>
>
>
>
>
>
> --
> Almudena Ponce-Salvatierra
> Macromolecular crystallography and Nucleic acid chemistry
> Max Planck Institute for Biophysical Chemistry
> Am Fassberg 11 37077 Göttingen
> Germany
>
|