Perhaps we could use Poincare's argument(?), that knowing one cross section of the universe in all of its detail would allow forward and back-calculation of all previous states. Then the universe would be its own lab notebook/ archive, and we would not need to bother with all of these technicalities in the first place. The images, then, could be back-calculated from the current (or any) configuration of all the universe's atoms, and then we could work better on improving our crystallography software (and ferreting out fraud) from those...

JPK

*******************************************
Jacob Pearson Keller
Northwestern University
Medical Scientist Training Program
Dallos Laboratory
F. Searle 1-240
2240 Campus Drive
Evanston IL 60208
lab: 847.491.2438
cel: 773.608.9185
email: j-kell...@northwestern.edu
*******************************************

----- Original Message ----- From: "Frank von Delft" <frank.vonde...@sgc.ox.ac.uk>
To: <CCP4BB@JISCMAIL.AC.UK>
Sent: Wednesday, March 18, 2009 1:19 PM
Subject: Re: [ccp4bb] images


Maybe, but images without experimental context (sequence? ligands? purification? crystallization format? -- PURPOSE OF EXPERIMENT!?!! relationship to the other 15 similar datasets) are as good as no images. And as far as I know, there's no good discussion on the table for that. At least, no-one on the thread mentioned it, so they're probably not thinking about it either.

I suppose efforts like PIMS or are a start, and maybe they can even have enough information (my feeling is they currently don't). But that's where the discussion should start: how to index (in sense of annotate) the datasets. The technicalities are just that: technicalities.

Or even closer to home: does ANY detector/beamline write even timestamps into the image header...? Never mind ring current, intensity of the beam, size of beam, size of crystal, length of direct beam path, etc etc...
phx



Gerard Bricogne wrote:
Dear Bernhard,

     I suppose you meant "ad nauseam" ;-) .
     In any case, what is the use of discussions and recommendations that
are not followed by action, and only result in making their contributors
themselves nauseated to the point of wanting to "put this to rest"?
     As Ethan has nicely stated in his reply to Garib's double-check of
whether we do need images, this matter should NOT be put to rest: it should be dealt with. As was argued at the end of the paper by Joosten, Womack et
al. (Acta Cryst. D65, 176-185), the main advantage of depositing images
would be that it would enable and stimulate the further developement and
testing of image integration and data processing software, to the same
degree that the deposition of structure factors has stimulated progress and
testing for structure refinement software.

Far from a boring issue only capable of giving headaches to Standards
Committee members, this is a vital issue: with each undeposited set of
images that contributed in one way or another to the determination or
refinement of a deposited structure, there disappears an opportunity to test
improvements in methods and software that would be likely to improve that
deposited entry (and most others) at a future stage. I think we need to take a long view on this, and abandon the picture of the PDB as a static archive
of frozen results: instead, it should be seen as a repository of what is
required not only to validate/authenticate the deposited models, but to feed
the continued improvement of the methods used - and hence, at the next
iteration, the constant revision and improvement of those very models. In
what way can this topic be a source of nausea?


     With best wishes,
     Gerard.

--
On Wed, Mar 18, 2009 at 10:16:42AM -0700, Bernhard Rupp wrote:

As Herb will attest, the need for keeping images and the various reasons
for it have been discussed ad nauseum and agreed upon in various imgCIF meetings - I am sure Herb or Andy Howard can provide links to the documents/recommendations, to put this to rest.
Best, BR

Past ACA Data Standards Committee serf
-----Original Message-----
From: CCP4 bulletin board [mailto:ccp...@jiscmail.ac.uk] On Behalf Of Kay
Diederichs
Sent: Wednesday, March 18, 2009 10:02 AM
To: CCP4BB@JISCMAIL.AC.UK
Subject: Re: [ccp4bb] images




Reply via email to