Dear Jeremy,
Thank you for the attached cartoon, most warmly welcome by all those in
need of a "displacement activity" in this gruesomely cold and rainy month of
April.
Oh those terrible French! I know them, I am one of them ;-) .
I found the Wikipedia entry on the subject quite entertaining: see
http://en.wikipedia.org/wiki/Discovery_of_Neptune
The conclusions in the "Later analysis" section will arouse suspicions that
it may have been written by a French author - however the graph given in the
previous ("Aftermath") section may be of interest, and speak for itself, in
our current likelihood-aware and (rightly) validation-obsessed frame of mind.
Back to serious things after this culpable diversion ... .
With best wishes,
Gerard.
--
On Thu, Apr 26, 2012 at 02:10:56PM +0900, Jeremy Tame wrote:
> The problem is it is not the PI who is jumping, it may be a postdoc he/she is throwing.
>
> Priority makes careers (look back at the Lavoisier/Priestly, Adams/LeVerrier or
> Cope/Marsh controversies), and the history of scientific reviewing is not all edifying.
>
> Too many checks, not enough balances. Science is probably better served if the
> author can publish without passing on the pdb model to a potentially unscrupulous
> reviewer, and if there are minor errors in the published paper then a competing
> group also has reason to publish its own view. The errors already have to evade the
> excellent validation tools we now have thanks to so many talented programmers,
> and proper figures and tables (plus validation report) should be enough for a review.
> The picture we have of haemoglobin is now much more accurate than the ones
> which came out decades ago, but those structures were very useful in the mean
> time. A requirement of resolution better than 2 Angstroms would probably stop poor
> models entering PDB, but I don't think it would serve science as a whole. Science
> is generally a self-correcting process, rather than a demand for perfection in every
> paper. Computer software follows a similar pattern - bug reports don't always invalidate the
> program.
>
> I have happily released data and coordinates via PDB before publication, even back in the
> 1990s when this was unfashionable, but would not do so if I felt it risked a postdoc
> failing to publish a key paper before competitors. It might be helpful if journals were
> more amenable to new structures of "solved" proteins as the biology often emerges
> from several models of different conformations or ligation states. But in a "publish or
> perish" world, authors need rights too. Reviewers do a necessary job, but there is a
> need for balance.
>
> The attached figure shows a French view of Le Verrier discovering Uranus, while
> Adams uses his telescope for a quite different purpose.
>
--
===============================================================
* *
* Gerard Bricogne [log in to unmask] *
* *
* Global Phasing Ltd. *
* Sheraton House, Castle Park Tel: +44-(0)1223-353033 *
* Cambridge CB3 0AX, UK Fax: +44-(0)1223-366889 *
* *
===============================================================
|