Dear all,
It has been quite fascinating to see this thread develop in the past
couple of days, as I was a hair's breadth away from initiating a similar
thread upon returning from the ACA meeting at the end of July.
An impromptu working dinner was held on the Tuesday evening of the
meeting to discuss various aspects of The Future of the PDB. Most of the
topics that were touched upon were technical, bordering (hardly) on the
clerical. I took advantage of a brief window of opportunity that opened
around the topic of "What should be the PDB's mission" to make a plea for
precisely the shift of emphasis that has been advocated collectively under
the "Importance of ..." thread:
(1) that people should be asked to deposit, and the PDB should archive,
raw images as well as all the information enabling the whole structure
determination and refinement process giving rise to a publication to be
reproduced by any interested third party; this would address the questions
of the reproducibility of results in a fairly radical (and beneficial)
manner;
(2) that the existence of such an archive would be enormously
beneficial to the software developers' community, as new developments could
be benchmarked against what was the "state of the art" at the time each
structure was solved, without the huge effort this involves at the moment;
(3) that the improvements in methods that such a working practice would
facilitate would themselves contribute to making it possible, in time, to
produce even better results from those annotated raw data than those
originally deposited; in this way, even the contents of the PDB would be
alive and constantly evolving, rather than frozen in their original state;
I was "surprised and disappointed" (standard euphemism) that the
obvious advantages of such an extension of the PDB's mission were met mostly
with reasons to not do it, with the expected arguments about the volume of
data etc ... . The fact that the PDB is giving its assent to the kind of
initiative that Ashley is talking about is mildly encouraging, but I concur
with others in thinking that this is too important to be left to volunteer
initiatives of this kind in the long run.
The side issue of verification and of spotting possible falsification
seems (as others have also mentioned) to be part of a bigger picture, which
is the risk of misbehaviour on the part of anybody who is put under
excessive pressure. Whatever the outcome of this particular incident may
eventually turn out to be, recent hiccups with structures published in
high-impact journals are a sign of a sickness in the system by which the
productivity of scientists is evaluated. We need to find ways of backing off
from this Hollywood-like fascination with (even, addiction to) these
journals, and from the pressure to "publish in Nature or Science, or
perish". I can remember Robert Huber telling me 20 years ago that we should
only publish in real journals, not in "magazines" (as he called Nature) -
and clearly, he had a point. A few years ago, Nature even started organising
conferences on the areas of science it considered as the hottest - a blatant
interference of mecantile media in the internal freedom of judgement of the
scientific community.
The two issues (a LivePDB, and the dictatorship of the high-impact
media) are clearly related, in the sense that a LivePDB would be a very
strong basis for calling to account the reviewers and editorial mechanisms
of these journals: this would occur "by default", instead of having to be
triggered by creating such traumatic "causes celebres" as that which emerged
last week.
With best wishes,
Gerard.
--
===============================================================
* *
* Gerard Bricogne [EMAIL PROTECTED] *
* *
* Global Phasing Ltd. *
* Sheraton House, Castle Park Tel: +44-(0)1223-353033 *
* Cambridge CB3 0AX, UK Fax: +44-(0)1223-366889 *
* *
===============================================================
