I would suggest that with 3.1 A data and only half of the scattering
mass known,
experimental phases would be an easier and better way to go than
refining the molrep solution.
Even sulphur phases could add if nicely measured, since you would
find the S easily
in the anomalous Fourier; but of course if its easy to do Se-Met so
much for the best.
Having said these, the 4-fold NCS might bail you out, but I would be
careful with mask
definitions in DM, it might need a bit more care.
It might also be an idea to switch the solvent correction off in REFMAC
(I am not sure, but I would try it):
SCALE ANIS TYPE LSSC SIMPLE (my syntax might be wrong but keywords
are correct)
SOLVENT NO
A.
On Feb 21, 2007, at 21:30, [log in to unmask] wrote:
> Dear all,
>
> Recently, I've gotten a 3.1A data set with a P1 space group. Based
> on a Mattchew
> coefficient, I have 4 hetero-dimer molecules in one asymmetric
> unit. Because one
> of the dimers was known already, I tried MR with a program Phaser
> and molrep
> (Amore doesn't calculate a translation function with a P1 space
> group). Phaser gave
> me a reasonable solution with Z-score closed to 20 and LL-gain over
> 900(molrep
> didn't give me a good solution), and that solution has 4 molecules
> with quite
> reasonable packing with local 2-fold symmetries. I also ran Phaser
> with
> different wavelength ranges with different truncated forms, and
> always the
> solutions were the same. However, when I calculated the map, the
> map was still
> messy and I couldn't see a clear density of the other molecules. I
> also ran a
> rigid body refinement and a restrained refinement with different
> programs
> (refmac and cns), but the R factor didn't drop. I ran DM with ncs
> averaging and
> solvent flattening, still the map wasn't improved. At this moment,
> we are still
> sure that our data set has a P1 space group with good statistics
> and the
> solution from phaser is quite reasonable.
>
> Is it possible that P1 space group can be a problem during MR? Do
> you think the
> solution from Phaser is just wrong or did I miss something?
>
> Any suggestions will be helpful to me.
>
> Thanks,
>
>
>
>
> uhnsoo
|