Re: [ccp4bb]: structure factors etc

[Phil Evans <pre@mrc-lmb.cam.ac.uk>]

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


The journals all require a Deposition code before publication. If that
was not issued until the SFs were deposited, that would police it

Providing data to the referees is another matter . . .

Phil



William Scott writes:
 > Hi Phil, Eleanor et al.
 > 
 > The main problem is this:  Who will do the compelling?  Currently each 
 > of the major journals has its own policy regarding deposition of 
 > coordinates, and having accidently put one to the test a couple of 
 > years ago while distracted with a sick baby, I can state that there is 
 > no apparent follow-up mechanism to check whether someone actually does 
 > what they say they will do.  (Maybe if someone else had caught it 
 > first, and complained, my experience would have been different.)
 > 
 > Perhaps it might therefore be more practical for us to tell the 
 > journals we won't referee a paper until we are provided with the data 
 > to do so (and perhaps also a letter signed by the authors agreeing to 
 > deposit the data).  In the case of MIR structures, I think at least the 
 > solvent-flattened phases should also be made available; perhaps even 
 > the initial MIR phases.  We all have learned that phases are at least 
 > as important as amplitudes, and MIR phases are still better quality 
 > control than even the best SigmaA maps.
 > 
 > Someone suggested Wilson plots, but these won't help much with a 3A 
 > structure, which presumably we are more worried about that one with 1.2 
 > A resolution and therefore more valid Wilson statistics.  Besides, if 
 > you have the Fobs, you can make your own Wilson plot.
 > 
 > I (perhaps regrettably) started this by trying to argue that with 
 > similar ideas, one could argue the need for open source code in 
 > crystallography.  This wasn't meant as an argument against depositing 
 > the data!
 > 
 > All the best,
 > 
 > 
 > Bill Scott