[ccp4bb]: Re: structure factors/quality control

[Thomas Stout <tstout@exelixis.com>]

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



Here's a heretical suggestion:  can we as the crystallographic
community who wish to have our "product" continue to be viewed
in high esteem (I presume), request that the PDB - perhaps through
the body of the IUCr - "certify" those structures which have
been deposited (and verified) with the supporting data, while
classifying those submitted as coordinates only as "preliminary"
or "unverified"?

That would shift the burden to the depositor to "do the right
thing" if they want their results to not be filtered from the
reliable portion of the PDB??

-Tom

(Who is now ducking for cover.....)



> From: "Ulrich K. Genick" <genick@brandeis.edu>
> Organization: Brandeis University
> Date: Mon, 25 Nov 2002 13:50:14 -0500
> To: ccp4bb@dl.ac.uk
> Subject: [ccp4bb]: Re: structure factors/quality control
> 
> ***  For details on how to be removed from this list visit the  ***
> ***          CCP4 home page http://www.ccp4.ac.uk         ***
> 
> The majority of the debate about the structure
> factors boils down to the integrity of the PDB
> data base and what we can do to keep it up
> or even improve it.
> 
> The usefullness of the PDB for crystallographers
> and non-crystallographers alike relies on the
> reliability of the structures deposited in it.
> 
> Bad structures, along with structure factors,
> which can be used to prove that those structures
> are bad, are only marginally better than
> bad structures alone. In particular non-
> crystallographers do not have the expertise
> or tools to use the structure factors any way.
> 
> Peer review, as frustrating as it can be, has
> proven to be the only viable quality-control
> tool in science.
> 
> And by peer review I do not
> mean a mob of crystallography vigilantes perusing
> the PDB to revel in their competitors (maybe honest)
> mistakes.
> 
> Here are some things we can all do to
> as our part in the peer reveiw quality control
> system:
> 
> - In our role as reviewers:
> WE HAVE TO INSIST THAT WE ARE
> PROVIDED WITH THE DATA, STATISTICS, FIGURES...
> TO ASSESS, IF THE CLAIMS OF THE ARTICLE ARE
> SUPPORTED BY THE DATA.
> Only a small fraction of the manuscripts
> I come across provides
> this level of information. We should just
> start to turn down papers which do not
> back up their claims. We should not
> assume "that they probably got it right".
> 
> - As authors:
> WE SHOULD PROVIDE THE MATERIAL THE REVIEWERS
> NEED TO MAKE AN INFORMED JUDGEMENT ABOUT OUR
> STRUCTURE.
> We should submit extra figures and
> tables/ statistics. If we are doing a good
> job during structure analyis, you will have
> the data at hand anyway. We just tell the journal
> we do not expect all this extra stuff to be published,
> we just want to make sure that the reviewers
> can check our claim.
> I am sure the editors will not complain.
> 
> - As teachers/PI's
> TRAINING AND QUALITY CONTROL IN OUR OWN LABS
> IS THE FIRST LINE OF DEFENSE AGAINST
> BAD STRUCTURES.
> From looking at many of the
> questions (and answers) posted on this bulletin
> board, I get the impression that many students
> do not understand the basic principles
> of refinement or handle details of refinement
> and structure validation
> with a certain degree of laissez-faire.
> Worse, apparently many people have
> nobody in their own lab who can or is willing
> to answer their questions on such very basic
> issues.
> 
> 
> -- 
> Ulrich K. Genick
> Assistant Professor
> Department of Biochemistry
> Brandeis University, MS009
> Waltham, MA, 02454
> 
> Room   Kosow 108
> Phone  781-736 2304
> Fax    781-736 2349
> Email  genick@brandeis.edu
>