These things are always very difficult.
It is obvious from the TRUNCATE log that your data is incomplete past
about 2.5A and that it is very anisotropic
(See wilson graph and the anisotropy fall off graph.) ( but why do you
say the data stops at 2.2 when the graphs go to 1.9A?)
There doesnt seem to be much sign of twinning to me - the acentric
moments are 3 instead of 2, which reflects the NCS but as you say it is
hard to be sure
The self rotation shows 422 fold symmetry and you could have a near
P4i2i2 space group.
Mt advice would be
1) try SFCHECK - its output is opaque but it takes anisotropy and
incomplete data and NCS translation into account as well as it can.
2) Try MOLREP with NCS and self rotation peaks taken into account
Eleanor
Julian Vivian wrote:
Hello CCP4BB,
I am having trouble with some data which I think is twinned
with pseudotranslational NCS. Any help would be greatly appreciated.
The crystals grew as very thin plates.
Running it through pointless suggests the data is P21 with cell dimensions
85.8 88.4 109.4 90.0 110.3 90.0
The data can be indexed as C222, yet will not merge.
There is likely to be pseudotranslation with a large peak in the native
patterson at 0 0 0.5
8 molecules in the asymmetric unit results in ~45% solvent.
We have a 3 wavelength SeMet dataset to ~2.2 Angs and an in-house native
to ~2.5 Angs. I
can't solve either by MAD/SAD/Molecular replacement with a 25% identical
homologue (though I think the structure will be highly conserved).
Britton analysis suggests the data aren't twinned (-h -k h+l), yet the
Yeates-twinning
server suggests twinning.
My fear is that the true twinning fraction is being hidden by the
pseudotranslation.
I would like some advice on how to estimate the twin fraction given the
pseudotranslation. In peoples' experience is there a chance that the
crystals are in fact perfectly twinned?
What are peoples' opinions on the best strategy for molecular
replacement in this situation?
Is there likely to be something other than twinning going on with this data?
So very many questions... Advice on any one of these points would be
very welcome.
For those that are curious, attached are the truncate log file,
self-rotation function and the output from the Yeates' twinning server.
Thanks in advance.
Regards,
Julian
------------------------------------------------------------------------
