look at the merging chi^2 value in HKL log files. If chi^2 is
systematically high (>2.0?) the high Rmerge could signal a problem with
detector or wrong space group. If the CHi^2 is around 1, assuming error
model is reasonable, it probably indicates weak data, which is quite
expectable with anisotropic data. This may or may not be compensated by
high redundancy (look at Rpim), but is not likely to be a problem, as
you have already solved the structure. If the annotator questions the
high R-merge, giving the chi^2 value should explain it.
Jarrod Mousa wrote:
> Hi all,
>
> I have solved the structure of a low resolution protein-protein complex
> to 4.5 angstroms (Rwork/Rfree=20/26%) in P42212, and I am trying to
> figure out if my data is good enough and where to cut off the
> resolution. I have processed the data in both hkl2000 and XDS, yet I get
> a high overall Rmerge and Rmeas, above 20-35% depending on the data set.
> The data is anisotropic, and I have tried the DAS server, but Rmerge
> remains high. XDS outputs in P4222, and I tried processing in P1 and
> putting in pointless and get P42212 as well. Phaser will take the P4222
> input and find a solution in P42212. Xtriage also recommends P42212. I
> am attaching two images from merging statistics, one cut off at the
> I/sigma >2, and the other cut off at 4.5 angstroms, as well as
> statistics after running through the DAS server at 4.5 angstroms.
>
> Just need to be sure my data is OK to publish, since I get a warning
> about high Rmerge during .pdb deposition. I am confident in the
> structure and the density looks good.
>
> thanks,
>
> Jarrod Mousa
>
|