 |
 |
Hi All..!!! Thanks for your excellent suggestions and comments. As suggested, I reprocessed the data (did not just expanded it...) in P1 space group (although the completeness was very bad, since we only collected 180 degree images) and ran MR in P1 with my latest model, after getting the single solution in phaser MR I did first cycle of rigid body and restrained refinement in Refmac5 and then fed that model with P1 mtz file to zanuda. Zanuda suggested P21212 to be the best space group (which phaser initially also suggested). However, the output model has a different origin compared to my initial P21212 phaser generated model. With this zanuda given model and the original P21212 mtz file, I ran rigid body refinement followed by restrained refinements (in the second cycle of restrained refinement, kept TLS, jelly body, local NCS refinement ON). Now the R/Rfree is slightly better compared to my initial phaser generated model (32/35 compared to 33/36). The model is not complete yet, I need to do further model building and refinements...will see how far the R/Rfree reduces...At least, now it is almost certain that I am dealing with the correct space group. I thank you all for your great help..!!! Best regards, Sudipta. On Thu, Jul 16, 2015 at 9:38 AM, Bernhard Rupp (Hofkristallrat a.D.) < [log in to unmask]> wrote: > >I think tNCS is indeed NCS, the problem is that molecules related by tNCS > are almost in the right position to be considered crystallographic symmetry > molecules, except that they are usually off by a small translation (I think > a small rotation could also provide this?) > > > > Yes, or a combination thereof, c.f. figures below > > > > >and therefore they are not crystallographic symmetry mates. > > > > Illustrated in Figures > > > http://www.ruppweb.org/Garland/gallery/Ch5/pages/Biomolecular_Crystallography_Fig_5-22.htm > > > > and > > > http://www.ruppweb.org/Garland/gallery/Ch11/pages/Biomolecular_Crystallography_Fig_11-07.htm > > > > But here is the problem: Looking closely at its related reciprocal space > figure > > > > > http://www.ruppweb.org/Garland/gallery/Ch11/pages/Biomolecular_Crystallography_Fig_11-08.htm > > > > you can see that the tNCS manifests itself in a commensurately modulated > superstructure, i.e. here > > the doubling of the cell resulting in weak lines of reflections between > the strong primary (if the NCS > > were actually crystallographic) smaller cell reflections. > > > > If the indexing software does ignore these and indexes a smaller cell > (i.e. misses the weak superstructure > > reflections) then there is nothing you can do to fix this. You need to > look at the actual images to find > > these weak, potentially missed reflections between the indexed spots, and > then reprocess with a > > correspondingly larger cell. Having only the smaller cell indexed, will > give effects similar to what has been > > described (somewhat similar to averaging). > > > > The same holds for incommensurate modulation (tNCS just being a case of > commensurate modulation) > > - looking at the diffraction pattern is necessary. > > > http://www.ruppweb.org/Garland/gallery/Ch8/pages/Biomolecular_Crystallography_Fig_8-30_PART1.htm > > > http://www.ruppweb.org/Garland/gallery/Ch8/pages/Biomolecular_Crystallography_Fig_8-30_PART2.htm > > > > BR > > > > > > I hope this helps, > > > > Tony > > > > ------------------------------------------------------ > > > > > > > *Dr. Antonio ArizaUniversity of OxfordSir William Dunn School of > PathologySouth Parks RoadOxfordOX1 3RE* > ------------------------------ >