[phenixbb] Only one solution in AutoMR with very negative LLG

Francis E Reyes Francis.Reyes at Colorado.EDU
Wed Mar 16 08:55:09 PDT 2011


> Could someone tell me why HKL2000 couldn't find the correct C2 SP?

  If you have a strong pseudotranslation (PST), you may see  
alternating weak/strong reflections along the axes. HKL2000 usually  
takes the most intense spots for indexing.Whether you include (or  
HKL2000 picks)  the weak reflections determines whether you get the  
large unit cell (only the strong reflections) or the smaller unit cell  
(strong and weak reflections).  LABELIT will work with PST quite well  
and I believe it picks up both the strong and weak reflections. It'll  
output a script suitable for mosflm. However, make sure you include  
images that include the strong/weak behavior (if it does exist in your  
images, b/c it can be xtal orientation dependent).

Watch the integration in P2, do you notice that some (weak)  
reflections (usually in between lattice lines) are missed?

> After refinement, the R went down to ~0.38, and the electron density  
> fits well with my model except some flexible parts.

Congrats on solving the phase problem.

If these flexible parts are important to your biochemical question  
(well as a crystallographer I would probably try to fit these parts),  
you should explore other space groups.

My suggestion: solve and refine in C2. At the end of this you'll find  
that [1] the structure is completely built, the density fabulous, but  
the R/Rfree is still high or [2] the structure is not completely  
built  (there's chain breaks, other abnormalities), or [3] something I  
haven't come across. Any of these will probably cause your structure  
to get rejected or at least disputed.

Take that structure and molrep into the P2 data using AUTO for the PST  
(pseudotranslation) vector (molrep will generate the second copy based  
on the PST vector found in the patterson and then use that for the TF  

Then take the solution and submit it to Zanuda (http://www.ysbl.york.ac.uk/YSBLPrograms/index.jsp 
)  to find the proper origin (in case molrep found the wrong origin).   
This program will test other (potential) space groups as well.

If you try this, I'd love to hear the results.


Francis E. Reyes M.Sc.
215 UCB
University of Colorado at Boulder

gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D

8AE2 F2F4 90F7 9640 28BC  686F 78FD 6669 67BA 8D5D

More information about the phenixbb mailing list