Yet Another DETWIN

[Kinkead Reiling <reiling@msg.ucsf.edu>]

***  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 have a mixed bag of NCS and Possible twinning and was interested in
input from the board.
The space group of my data is P3121, but the data will merge in P622
with comparable Rsyms
A molecular replacement of my dimer in P3121 reveals that the NCS two
folds lies along a P6122 two fold at fractional height 1/12.

I have tried refining in P3121 and the R-Free hangs at 40%

Running my data on the Yeates twin server I get back
 twin fraction of 0.440387 with and operator of -h -k l

Thus the following DETWIN output
*********************************************************************************

Correlation coefficient (on intensities) for input data 0.986
 Correlation coefficient after detwinning 0.348

 Number of reflections written to output file  13618
 Number omitted because the twin mate was not observed    149

  Correlation coefficient as a function of resolution.
    Dmin   Number    CC before     CC after
                   detwinning    detwinning
    8.67      309     0.974         0.036
    6.20      836     0.981         0.185
    5.08     1090     0.983         0.252
    4.41     1230     0.976         0.085
    3.95     1390     0.983         0.259
    3.60     1507     0.982         0.218
    3.34     1672     0.983         0.250
    3.12     1771     0.980         0.179
    2.95     1875     0.969         -.047
    2.80     1938     0.951         -.272
**********************************************************************************

So the question is what are the best ways to separate NCS symmetry in
ones data from something such as twinning?
And could the NCS be making the apparent twin fraction higher and thus
the non-zero detwinned CC's?

cheers and thanks for the support
kinkead reiling
UCSF