This seems like a classic twinning case.
The plots of the cumulative intensity distribution and the moments shown
after TRUNCATE should alert you to this well before attempting any
structure solution..
As Jim Warren says, you can refine even a 50% twin with CNS or SHELXL
You will need to select the correct space group.
Look at the program documentation - general remarks on twinning...
It points out that for a true point group PG6 2 2 no twinning is possible
*P6i22*:(h,k,l) already equivalent to (-h,-k,l) /and/ (k,h,-l) /and/
(-k,-h,-l) so no twinning possible.
However a perfect twin for the Laue group ,P312, P321 or P6 might appear
to have this system.
But since you have P61 xx symmetry I guess the only possible true case
is P61 with twinning operator k,h,-l
Eleanor
Sue Roberts wrote:
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Hello Everyone
I'll post a summary after I've waded through the (many) answers, but I
seem to have left out some information. There are some interesting
suggestions.
I've already checked enantiomorphic, and all the other P6x22 and P6x
space groups. Only P6122 gives a good MR solution (well, so does P61,
P31, P312, but these MR solutions have the same packing as P6122. MR
solutions in lower symmetry space groups with fewer molecules/AU have
low Z-scores, and high R (when I try to refine it), and maps that tell
me to add more molecules.) (I've also gone as low in symmetry as P31
(etc) without any improvement in the situation. Heck, I've gone to
orthorhombic symmetry too.) After the MR solution in P6122, I could
see clear density for helices and side chains that were not in the
model structure. (I really like phaser because it gives maps.)
The model I used is only about 60% of my protein, consequently, the MR
solution has no packing problems. It is in building the remaining
part of the structure than I run into the problem. R/Rf are 0.27 /
0.33 for 2.3 A data. At this point, there is nothing more to build,
the maps are flat. I have no information as to what the structure of
those 30 residues are - they're not in the model structure I used.
Since some of these are implicated as catalytically important
residues, I'd really like to see them. (I know, grow crystals under
different conditions).
Interestingly, if I download the structure factors for the structure I
used for the MR (a thanks in general to all who deposit structure
factors), the structure I used as a model seems to have the same
problem. A big chunk of the protein is missing and the molecule runs
into itself across a two-fold (it's a different cell and space group).
Sue
begin:vcard
fn:Eleanor Dodson
n:Dodson;Eleanor
email;internet:[EMAIL PROTECTED]
tel;work:+44 (0) 1904 328259
tel;fax:+44 (0) 1904 328266
tel;home:+44 (0) 1904 424449
version:2.1
end:vcard
