A Plea To John K! (was Re: Comment to S.Blocklehurst)

David Jones jones at bsm.bioc.ucl.ac.uk
Sat May 6 08:50:58 EST 1995

Peter Sibbald (sibbald at qucis.queensu.ca) wrote:

> Just a quick aside here... The rumour as i understand it is that George
> Rose has made major progress on the protein folding problem and is
> publishing it in _Proteins_??? Not Nature? Either this is not a solution
> to the protein folding problem or i am not the only one to think that
> Nature has slipped a little in past years. ;-)

There are all sorts of unknowns here. Perhaps George Rose thought Proteins
was more appropriate? Perhaps the method is very complex and couldn't be
slotted into a short publication? Perhaps George Rose decided to err on
the side of caution and submit to a specialist journal rather than one
with a general readership?

Of course he might have submitted it to Science or Nature and had it
rejected? That's not to say that the paper is necessarily flawed, but
perhaps the referees thought that whilst the results were good, the
method was not actually new, but that George Rose had managed to find a
very good recipe for his potentials. This must remain as speculation
until we know more of course.

One point to make, which sort of relates to my very first comment, though
originally made rather cryptically - is how do you prove to someone that
you have solved or almost-solved the problem - indeed, how do _you know_
you have it cracked? My feeling is that Nature and Science will treat
solution-to-the-protein-folding-problem papers only slightly differently
from perpetual-motion papers. The difference being that perpetual motion
papers probably won't get to referees and protein folding papers probably
will. However, and this is of course conjecture, they are likely to
magnify the negative referee comments much more that they would
in some other cases. Any statement like "good but it still remains to
be seen..." or "seems to work, but" in the referee reports are likely
to result in immediate rejection. I guess it is understandable in that
Nature and Science do not want to damage their reputation by publishing
something which later proves not to live up to the initial promise.

This is why meetings such as the recent "Meeting on the Critical assessment
of Techniques for Protein Structure Prediction" at Asilomar have been
organized recently. Although it is probably the case that George Rose
didn't have his stuff running in time for the submission deadlines for
the 94 meeting, these blind-test results are generally considered the most
convincing. Steve Benner and some other groups of course publish one-off
blind predictions over the period of a year or so, and this is another
route to convincing critics. Of course people will always be skeptical
until they have convinced themselves by applying it sucessfully themselves
for example - but that's the name of the game in this field.

If George Rose's stuff proves as good as it sounds [or is rumoured
to sound :-)] then perhaps it can be tried out in the prediction
course that was announced recently (and in which for my sins I've
agreed to be one of the instructors). Of course I mention that in
an entirely unofficial off-the-cuff capacity and this might well not
agree with the feelings of the organizers or my fellow instructors.
Don't you just love all these disclaimers we all feel it necessary to
insert at regular intervals these days! ;-)

This message was written, produced and executively directed by Dr David Jones
Address: Department of Biochemistry and |     Email: jones at bsm.bioc.ucl.ac.uk
Molecular Biology, University College,  |       Tel: +44 171 387 7050 x3879
Gower Street, London WC1E 6BT, U.K.     |       Fax: +44 171 380 7193

More information about the Proteins mailing list