Retraction Watch

Tracking retractions as a window into the scientific process

Investigations into Voinnet’s work announced; critic publishes original peer review

with 18 comments

Olivier Voinnet

Olivier Voinnet

The story about Olivier Voinnet, a high-profile plant biologist whose work has fallen under scrutiny, continues to build momentum. Late last week, Voinnet’s employer and one of his funders announced they were investigating his work, and one of the peer reviewers of a soon-to-be-retracted paper has made her original report public.

Here’s the statement from CNRS, which funds Voinnet:

Over the past few months, a number of anonymous comments have been posted on the PubPeer website to report diagram/chart manipulations concerning around thirty articles signed or co-signed by Olivier Voinnet, CNRS senior researcher currently on secondment at the Swiss Federal Institute of Technology (ETH) in Zurich.

Faced with these allegations, the CNRS set up a commission of inquiry, made up of leading experts. In a context where the works of a significant number of researchers have been massively and anonymously called into question, public institutions have a duty to act in strict compliance with legal and ethical standards. These standards do not allow any public statement to be issued prior to completion of the procedure, in order to ensure that an in-depth analysis of the situation is carried out, in which all parties can freely express their views. The CNRS abides by these rules, which guarantee rigor and fundamental protection of the rights of individuals. The organization will of course face its responsibilities.

Irrespective of the works of this commission, the CNRS notes at this stage that these public allegations referred to the presentation of certain charts/diagrams but that, to its knowledge, no declaration has challenged the overall results obtained by Olivier Voinnet and his colleagues on the role of small RNAs in the regulation of gene expression and antiviral response — these results having been confirmed on several occasions, whether using the same or other material, by various teams worldwide.

And here’s a statement from ETH Zurich:

Allegations that were made online against an ETH professor have now been picked up in the media. On learning of the accusations, ETH Zurich immediately appointed an investigative commission.

ETH Zurich became aware in January of internet comments expressing the suspicion that images may have been manipulated in publications. Olivier Voinnet, Professor of RNA Biology at ETH Zurich, is listed as author or co-author of these publications. The allegations, published on the websites PubPeer and Retraction Watch, concern the illustrations in these publications; the studies’ findings are not in doubt.

ETH Zurich takes these allegations very seriously. The university’s Executive Board instructed Professor Detlef Günther, ETH Zurich Vice President for Research and Corporate Relations, to immediately set up an investigative commission tasked with conducting a judicious and impartial assessment of the situation. The investigation will be led by a four-person commission and chaired by an external expert independent of the ETH Executive Board. All relevant publications will be examined. Since the investigation will be comprehensive, it is not yet clear when the commission will deliver its findings. ETH Zurich will give a full report as soon as all the facts have been satisfactorily established. ETH Zurich has also established contact with the French National Center for Scientific Research (CNRS) and the University of Cambridge. Both of these institutions are also named in these allegations and have set up their own independent investigative commissions.

ETH Zurich has a Procedure to address allegations of research misconduct at the ETH Zurich that applies in this case. The investigation will adhere strictly to this defined process. The investigative commission will submit its report to the Executive Board, which will then determine, on the basis of the facts presented, the next steps and any required measures.

“These allegations have come as a surprise to the Executive Board at ETH Zurich. Olivier Voinnet is a scientist whose outstanding research findings have been confirmed repeatedly by other research groups,” says Günther. “In order to make an assessment of the allegations, we must await the findings of the investigative commission. It is now of the utmost importance – both for ETH and for Olivier Voinnet – that we allow the investigative commission the time and space they need to carry out their work and deliver an independent evaluation. We will therefore wait until we have received the commission’s report before issuing any further comment.”

We found some of the language in the announcements puzzling. Call us old-fashioned, but generally it’s a good idea to actually do an investigation before saying that “the studies’ findings are not in doubt.”

Apparently, Vicki Vance, who raised the red flag more than a decade ago about a to-be-retracted paper by Voinnet and colleagues in Plant Cell, also had some concerns. Vance has posted her original peer review of the manuscript at ResearchGate, and has also sent an open letter to both institutions. An excerpt:

…I have read that the posts showing fabrication of data in the figures of many of Prof. Voinnet’s articles were viewed by some people as having little importance.  The rationale being provided is that the results are still valid because other labs have been able to show the same results. That is NOT completely true. The practice of fabrication of data by the Voinnet lab has had serious negative impact on the field of RNA silencing. Many investigators are, in fact, not able to repeat some aspects of his reported results or have conflicting data. However, once results are published in high impact journals by a powerful and important senior investigator such as Prof. Voinnet, there is little chance to get funding to pursue conflicting data and further experimental approaches are stalled.  I am a tenured scientist, approaching academic retirement and can therefore afford to bring these items to light.  However, because the consequences of Prof. Voinnet’s unethical behavior are not yet clear, many scientists more junior than I am are afraid to speak up and risk the wrath of Prof. Voinnet, should he remain the powerful force he has been in the past.

In summary, I think that Prof. Voinnet’s unethical behavior has damaged the field immensely because it is no longer clear what is true in his work and what is fabricated.  In my opinion, these are serious incidences of scientific misconduct and I hope that your investigations will consider them as such.

We’ve made Vance’s entire letter available here.

The European Molecular Biology Organization (EMBO), of which Voinnet is a member, is awaiting the outcomes of the institutional investigations before considering next steps.

Like Retraction Watch? Consider supporting our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, and sign up on our homepage for an email every time there’s a new post.

Written by Ivan Oransky

April 13th, 2015 at 2:00 pm

Comments
  • Clarity April 13, 2015 at 2:15 pm

    Dr. Vance states: “fabrication of data in the figures of many of Prof. Voinnet’s articles”
    Emphasis on the plural, articles, and thus not referring only to The Plant Cell paper.

    • Bobo April 13, 2015 at 4:04 pm

      That quote is based on her assessment of PubPeer postings, not her own experience.

  • Leonid Schneider April 13, 2015 at 2:18 pm

    For more background regarding the retraction announcement at The Plant Cell, here is my report for Laborjournal/Lab Times: http://www.labtimes.org/editorial/e_600.lasso

    • herr doktor bimler April 13, 2015 at 6:33 pm

      Leonid writes”
      Enter “The Third Referee”

      “It’s always the third goddamn reviewer that screws us over!!”

  • Didier Torny April 13, 2015 at 5:09 pm

    Vicki Vance is right about the “dead ends” that could result from data fabrication, leading other teams to unsuccessfully build on “empty science”. Nevertheless, the sole fact that some results have been “reproduced” does not indicate that the first experiment was well performed, it rathers shows the real ability of a dedicated scientist to predict a result based on current knowledge and some extrapolation :

    “If the absence of reproducibility is often considered a clue to falsification, the opposite is not necessarily true: a highly skilled forger can produce a sort of retroplagiarism, arranging or fabricating data that others, through slow scientific work, will have actually produced and analyzed.”
    https://hal-mines-paristech.archives-ouvertes.fr/hal-00780561/document

    PS : this comment is an epistemological one, whether Voinnet’s experiments were authentic or not.

  • Morty April 13, 2015 at 6:07 pm

    CNRS’s decision to set up a independent commission of inquiry made up of leading experts is very welcome, and should stand as a good example for other institutions like MD Anderson and Wayne State, which
    also have some well known cases.

  • Miguel Roig April 13, 2015 at 6:08 pm

    I agree with Didier’s quote, though I strongly suspect that the level of skill needed to fabricate numerical data (as opposed to images or other more complex data) has more to do with the fraudster’s familiarity with actual data outcomes than with any other subversive skill. For example, if someone who has done research with, say, the Stroop effect (Psychology) wanted to fabricate human reaction time performance data for that task, s/he could easily do so based on a thorough familiarity with how such data are typically produced by subjects under various conditions. Thus, assuming the data in question are in the form of seconds/milliseconds, a fraudster might simply generate a desired pattern of results in seconds while insuring that the last couple of digits after the decimal in each data point are truly random. After all, humans are known to be incapable of producing random strings of digits and that lack of randomicity at the millisecond level would be a dead give-away.

  • Leonid Schneider April 14, 2015 at 1:28 am

    I hope for the simple explanation that those statements in the ETH Zürich and CNRS press releases about only the “Illustrations”, but not the research and data being questioned are only bad formulations by ill-advised press communication officers.
    Moreover, ETH investigations were supposed to end April/May 2015, as I was officially told. I don’t know if it’s a sign of progress or of delaying, that they will now take much longer.

  • Miguel Roig April 14, 2015 at 6:23 am

    And for those who might not be familiar with the quote:

    https://www.youtube.com/watch?v=n6jfVR98M-E

  • Raphael Levy April 15, 2015 at 9:45 am

    My take on this PR with an analogy to the possible response of another public institution (the police) to serious accusations against a senior member of staff: https://raphazlab.wordpress.com/2015/04/15/whats-wrong-with-that-cnrs-press-release/

  • Impact April 16, 2015 at 5:15 pm

    In addition to the PNAS 2012 correction:
    http://www.pnas.org/content/early/2015/04/15/1506636112.extract
    Can anybody please provide the Altmetric score/data of OV, using the Altmetric Explorer:
    http://www.altmetric.com/aboutexplorer.php

  • Leonid Schneider April 16, 2015 at 5:48 pm

    In regard to the Brodersen et al 2012 PNAS correction: Someone made a comparison of Fig 5C original vs corrected.http://i.imgur.com/PxuvEbw.jpg
    Authors declare they only fixed loading controls, yet they replaced more that the single duplicated band originally criticised (http://imgur.com/VwExhkV), but also exchanged a panel which was not loading control at all (AGO1 ECL)

    • Leonid Schneider April 17, 2015 at 1:04 am

      A simple observation: originally someone pointed out, one single band image was duplicated in Fig 5C. Nothing else. Authors had to correct much much more in the Fig 5C instead. Hope they and PNAS checked other figures as well.

  • J April 16, 2015 at 7:57 pm

    Ah, I’m not convinced they have. The ‘new’ AGO1 ECL figure *could* be the old one, but squashed and contrast adjusted? Hard to tell with the resolution given, not that I think it matters much at this stage.

  • PNAS April 17, 2015 at 5:03 am

    From PubPeer, related to the 2012 PNAS paper:
    https://pubpeer.com/publications/FD08039FB188E781A80C1C469FFFAB#fb28566
    This paper was originally submitted through the highly controversial “Prearranged Editor” option of PNAS (https://www.insidehighered.com/news/2009/09/14/pnas or http://www.nature.com/news/scientific-publishing-the-inside-track-1.15424).
    Here, the author is free to invite/appoint any NAS-member to handle his/her submission. This pre-arranged editor also leads or is directly involved into the peer review process. In this case, the prearranged editor was James Carrington, with whom Voinnet has published two Science papers some years before this submission:
    https://pubpeer.com/publications/16741077
    https://pubpeer.com/publications/20413458

    Perhaps Dr. Carrington would care to comment?

    Editor’s note: Carrington did comment, and corrected an error in this comment. Please see Leonid Schneider’s reply to this comment.

  • Leonid Schneider April 19, 2015 at 6:40 am

    Some updates:

    PNAS paper by Navarro & Voinnet from 2013 is criticised for neglecting to deposit replicates of central data on RNAseq https://pubpeer.com/publications/0D3854D679A48F32024A542C6679E2

    Another Voinnet paper now directly criticised on PubPeer for data inconsistencies (Gibbings et al, Nature Cell Biology 2012) https://pubpeer.com/publications/E2451CF219109D92E4BFAD59410425#fb28598 J

    James Carrington explains the peer review process behind the recently corrected Voinnet paper in PNAS, Brodersen et al 2012, and corrects an error in the comment above https://pubpeer.com/publications/FD08039FB188E781A80C1C469FFFAB#fb28615

  • Leonid Schneider May 4, 2015 at 5:06 pm

    Another update:
    Apparently, Olivier Voinnet was confirmed as invited speaker at the “Microsymposium on Small RNAs” organized by IMBA (Institute of Molecular Biotechnology of the Austrian Academy of Sciences) and IMP (Research Institute of Molecular Pathology) in Vienna.
    http://www.imba.oeaw.ac.at/microsymposium
    However, he did not participate as I was informed by the organisers. This cannot be a short notice cancellation, since while Voinnet is named on the conference poster (http://www.imba.oeaw.ac.at/typo3temp/pics/p_9aa11339f6.jpg), he is not listed as speaker at all on the conference booklet: http://www.imba.oeaw.ac.at/fileadmin/imba_neu/documents/Symposia/Program_Microsymp15.pdf

  • Post a comment

    Threaded commenting powered by interconnect/it code.