Re: [ccp4bb]: NCS averaging

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

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

Dear Jinsong,

If you have translational NCS of the type you suggest, you should
see a big peak in the native Patterson map at (0,0.1,0.5).  Do you
see that?

The bad news is that, if this is the case, NCS averaging is not
going to help you as much as you might hope.  NCS averaging gets
its power from the relationships between structure factors created
by molecules in different orientations, but your molecules are in
the same orientation.  Apologies for advertising, but Gerard
Kleywegt and I discuss this in a review article we published in 
Structure a couple of years ago.  You might find that interesting, 
because it also discusses the question of how to define the NCS 
operators and envelope from the electron density map.


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

On Thu, 05 Oct 2000 09:10:46 -0700, Jinsong Liu <jsliu@tularik.com> wrote:
>
> ***  For details on how to be removed from this list visit the  ***
> ***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***
>
> Thanks for so many replies. I think I need to explain what I have done
> already.
>
> Before this MAD  (2.2A) data set, we have a 2.4A native dataset and I have
>
> tried MOLREP with a poor MR search model. The self rotation function did
> not
> show a proper two-fold.  After many trials in AMORE, MOLREP, all the
> solutions
> suggest that two molecules have the same rotation but with translation
> shift by half
> in Z (and less than 1/10 in y). This is in P21.  But the MR solutions is
> not refinable.
> so when I collected the MAD data, I tried GETAX after phased by SHARP, and
> it
> didn't work due to lack of proper 2-fold.
> So that is why I want to find the better way to do NCS-averaging.
> I will try use the MR solution to get a NCS mask and doube-check if the MR
>  solution
> is at least in the right place compared with the MAD map. And I will
> consider so many
> sugestions your guys offer and post a summary later.
>
> Thanks again.
>
> Jinsong
>