The PDB is already taking action on this question by setting up
validation task forces for X-ray and NMR structures. I'm chairing the
X-ray task force, which is finally nearing completion of its report
after working on it since April 2008.
One of our recommendations (of most relevance to this discussion) is
that, when a structure is deposited, a validation report should be
generated in the form of a PDF file, including an intuitive
representation of overall quality that should be understandable even
by non-crystallographers. Editors and referees would know that this
report was available and could (should) request it; eventually
journals could require it, in the same way that they now require PDB
codes. Probably the best way forward is for professional bodies such
as the IUCr to consider this as policy, which they could recommend to
the journals.
Credit where credit is due: George Sheldrick originally proposed such
a mechanism, as part of the previous discussion on "using our
validation tools".
Although improved validation tools (especially ones using the newly-
required diffraction data) will make it harder to commit convincing
fraud, someone determined to do it will find a way to satisfy whatever
tools exist at the time; it will only be tools developed later that
will catch them. But fraud is very rare, and my personal opinion is
that it would be counter-productive to be so suspicious of our
colleagues that we spend our time running forensic checks on every
result. Far more damage is done by well-intentioned incompetence, and
that's the main purpose of validation.
Regards,
Randy Read
On 11 Dec 2009, at 09:30, Prof. Joel L. Sussman wrote:
11-Dec-2009 11:30 Rehovot
Dear All,
I Agree fully with Tommi, and feel, in parallel, we in the MX
community must think of better tools for referees to review papers
and insist that these be followed. For example we should insist on
getting BOTH the coords and structure factors for papers submitted,
so it would be possible to do test on the structures which appear to
have 'question marks'. Without this data, it is very difficult to do
a serious job in refereeing.
Also, any manuscript, submitted to any journal, where BOTH the
structure factors and coords are not submitted to the PDB should be
rejected immediately without further review. This should be the
policy of the reviewer, independent of any particular policy of the
journals.
It might be worthwhile for the MX community to also consider writing
a strong letter to the journals in which published the papers for
which the structures have now been retracted, to state just how
serious this matter is and that the journals should also take much
more responsibility of finding better ways to have these papers and
structures reviewed.
Joel
-------------------------------------------------
Prof. Joel L. Sussman
Director, Israel Structural Proteomics Center
Dept. of Structural Biology
Weizmann Institute of Science
Rehovot 76100 ISRAEL
Tel: +972 8-934 4531
Fax: +972 8-934 6312
joel.suss...@weizmann.ac.il
www.weizmann.ac.il/~joel; www.weizmann.ac.il/ISPC
-------------------------------------------------
On 11 Dec 2009, at 11:19, Tommi Kajander wrote:
Would the exact analysis of how each of these things were wrong and
fabricated be somewhere
available???? Would be fair (apart from the known case of C3b) to
have the whole analysis available
instead of just this kind of news feed. I suspect its not obvious
by five minute check in all cases.
Perhaps there needs to be ways within PDB in form of automated
tools that would raise those red
flags in suspicious cases (e.g. some data analysis --such as the
contribution by solvent etc now that data beyond 8Å
is by default used in refinement) - as it appears peer review/
editing by journals isn't/cant always be(?) stringent enough.
In any case, some type of automated analysis of the whole data
base might be a good idea, as there can be
other cases (with another couple of thousand papers citing them..).
tommi
On Dec 10, 2009, at 4:16 PM, Ibrahim Moustafa wrote:
"After a thorough examination of the available data, which
included a
re-analysis of each structure alleged to have been fabricated, the
committee
found a preponderance of evidence that structures 1BEF, 1CMW,
1DF9/2QID,
1G40, 1G44, 1L6L, 2OU1, 1RID, 1Y8E, 2A01, and 2HR0 were more
likely than not
falsified and/or fabricated and recommended that they be removed
from the
public record," the university said in its statement this week."
------
Randy J. Read
Department of Haematology, University of Cambridge
Cambridge Institute for Medical Research Tel: + 44 1223 336500
Wellcome Trust/MRC Building Fax: + 44 1223 336827
Hills Road E-mail: rj...@cam.ac.uk
Cambridge CB2 0XY, U.K. www-
structmed.cimr.cam.ac.uk
