The 34-34-34 cell does not predict all the spots, does it?
from the diffraction pattern it seems only the 34-34-170 or 34-34-340  
cell can predict all spots, so the structure should be solved in the  
one that predicts all spots.
The procedure I would use is to take a 180º dataset, sacrificing some  
resolution if necessary, then integrate in P1 and solve the structure  
by MR or if possible by just placing the double helices in the cell.
If you then still have disorder of the bases this means that the dsRNA  
can fit in the density in several ways, either up and down, or, less  
likely, if the ends of the 10 bp duplexes are not clear, even  
displaced by one of more bases along the long axis. In principle there  
is no reason why the differences in sequence ("inside" structure)  
always have to lead to significant differences in the "outside"  
structure (the phosphates) and thus affect the crystal packing. They  
usually do, but not always (some dimeric transcription factors  
crystallised with asymmetric dsDNA duplexes have disordered bases, as  
long as the bases with which the protein interacts are the same the  
protein doesn't necessarily care that much about the rest...).
Once you have your structure clear in P1, you can introduce the  
possible symmetry axes one by one; and once you have your spacegroup  
clear, collect as high as possible resolution data with the minimal  
wedge necessary and as high a dose as possible without inducing  
radation damage.
If your duplexes fit in two ways, "up" and "down", refinement should  
be possible with both at partial occupancy.

Mark J. van Raaij
Dpto de Bioquímica, Facultad de Farmacia
Universidad de Santiago
15782 Santiago de Compostela
Spain
http://web.usc.es/~vanraaij/







On 29 Jan 2009, at 10:45, Margriet Ovaere wrote:

> Dear all,
>
>
> There were some comments about detector issues, but these can be  
> ruled out, to my opinion, since the lines appeared on different  
> beamlines.
>
>
> Default settings of mosflm (spot picking) finds the cell 34 34 34 90  
> 90 90 (pointless indicating P41212)
>
> Structure was solved by SAD phasing on the phosphates in this space  
> group. Double helices stack in continuous helices, the backbone is  
> well defined in the (refined) density maps but the individual bases  
> are messy (purines and pyrimidines seemed to overlap) + obviously  
> not all spots were covered and the duplex does not fit in the A.U.
>
> For this reason the integration was repeated in the higher cell 34  
> 34 170
> Space group most probably P212121, but solutions can be found in  
> P41212 as well (still disordered bases)
>
> There are also indications that the 41 screw axis is rather a pseudo  
> axis than a pure crystallographic one, also in the small cell
>
> Reindexing the cell to 34 34 340 also gives a solution, which  
> supports the theory of Holton
>
> Rmerg is around 5% for the small cell, about 8% for the 170Å cell  
> (both in P41212)
>
>
>
> Which refinement procedure would be best to follow?
>
> kind regards
>
> Margriet
>
>
>
>
> Margriet Ovaere
> Chemistry Department K.U.Leuven
> Biomolecular Architecture
> Celestijnenlaan 200 F
> B-3001 Heverlee (Leuven)
> Tel: +32(0)16327477
>
>
>
>
>
> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm for  
> more information.
>