[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ccp4bb]: Twinning/Merging

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

Alright guys,

>I disagree too, like Jim, and would not use Rmerge = 16% data for
>refinement, but rather transfer them into Trash...
>Rudolf Ladenstein

Rudi: I am sure you have had datasets with that kind of Rmerge in the 
past and found that they were usable... Sometimes, you have to live 
with what you've got.

To go back to the original question: Daniel, to confirm twinning a 
possible source for the "Rmerge misbehavior", you would need to do a 
twinning analysis for each of the three individual batches. You 
didn't give a lot of information, but what is the Rmerge statistics 
for each individual batch? Is the Rmerge fine for each batch and goes 
up only after merging the batches? How does the cumulative intensity 
distribution look like (for each batch)? How does the I/sigma(I) 
statistics look like? Is there anomalous signal in the data? Is there 
anisotropy in the data? Have the batches been indexed in the same 
setting, etc. etc.? We will need to know a lot more to answer the 
original question.

Finally, to come back again to my utter disrespect for Rmerge 
statistics, I would like to know what the quality of the electron 
density from the room temperature dataset was and if it was 
sufficient to see the substrate.

Long live high Rmerge!

Mischa Machius, PhD
Assistant Professor
University of Texas Southwestern Medical Center
Department of Biochemistry
5323 Harry Hines Blvd.                                Tel: +1 214 648 9760
L4.250                                                Fax: +1 214 648 8954
Dallas, TX 75390-9038, U.S.A.     Email: Mischa.Machius@UTSouthwestern.edu