Re: [ccp4bb]: Questions about translational NCS

[Randy Read <rjr27@cam.ac.uk>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

On Thursday 20 February 2003 02:16, Jianghai Zhu wrote:
> ***  For details on how to be removed from this list visit the  ***
> ***          CCP4 home page http://www.ccp4.ac.uk         ***
>
> Hi, all,
>
> I have several questions about translational NCS.
>
> 1) If I have a pure translational NCS, I can find a peak in my
> native patterson map. But if I have a NCS happen to be parallel
> to my crystallographic symmetry, they generate some
> translational vectors and I can find a peak in my native
> patterson map too. How do I know it is a pure translational NCS
> or a NCS parallel to the crystallographic symmetry?

If an NCS n-fold rotational axis is exactly parallel to a crystallographic 
axis that shares the factor n (e.g. 2-fold parallel to 2, 4 or 6-fold; 3-fold 
parallel to 3-fold or 6-fold), the combination of NCS and crystallographic 
symmetry will generate a pure translational NCS.  If the axes are both pure 
rotational axes, the translation will be in the plane perpendicular to the 
axes.  Any screw components will give an out-of-plane contribution to the 
translation vector.  What you choose to call it is partly a question of 
semantics.  If I had a big Patterson peak at u,v,1/2 in P2(1)2(1)2(1), I 
would interpret that as a dimer with a pure 2-fold parallel to the z axis, 
centered on a position shifted by u/2,v/2 relative to one of the screw axes 
parallel to z, but would point out that this generates translational NCS.  On 
the other hand, if the Patterson peak was at u,v,w where none of u, v or w 
was equal to 1/2, I would interpret that as translational NCS.  If I wanted 
to, I could interpret this instead as showing a dimer with a local screw axis 
parallel to either x, y, or z.  But an assembly with point group symmetry 
makes more sense as an object possessing non-crystallographic symmetry.  

>
> 2) I was told that the translational NCS could make the data
> look like a data from a twinning crystal. Is that true? If it
> is, could somebody explain it to me or point out some
> reference? Will the pure translational NCS and NCS parallel to
> crystallographic symmetry give the similar intensity
> distribution?

Actually, translational NCS modulates the diffraction pattern to make some 
reflections systematically more intense (where the translationally-related 
molecules scatter in phase) and some less intense (out of phase), so the 
distribution of intensities is broader.  Twinning has the opposite effect, 
narrowing the distribution of intensities.  So, if anything, the presence of 
translational NCS can mask twinning in the common twinning tests. 

This situation, where an NCS axis is nearly parallel to a crystallographic 
axis, is actually quite common.  There's a paper by Wang & Janin (Acta Cryst. 
D49:505-512, 1993) showing that this happens frequently in structures 
deposited at that point in the PDB.  It may even be more common than their 
data suggest, because there was a time when people threw away crystals with 
NCS to reduce the size of the data collection and computing problems.

-- 

Randy J. Read
Department of Haematology, University of Cambridge
Cambridge Institute for Medical Research      Tel: + 44 1223 336500
Wellcome Trust/MRC Building                   Fax: + 44 1223 336827
Hills Road                                    E-mail: rjr27@cam.ac.uk
Cambridge CB2 2XY, U.K.                       www-structmed.cimr.cam.ac.uk