Olivier Voinnet, the new Dreyfus?

Version française ici

Much of French media and academia, and certainly also the international plant science community now debates a hot conspiracy theory: what if Olivier Voinnet is actually innocent, a visionary genius who fell prey to a conspiracy of fraudulent colleagues and scheming bureaucrats? Indeed, French research elites and even the Government of the French Republic now raised an accusing finger at the real culprit of all French research misconduct scandals: Leonid Schneider. I tend to agree!

The Count of Monte Cristo

In 2015, Voinnet was found guilty of research misconduct, after data manipulations were reported in around 40 of his papers, virtually the entire publication record from his PhD thesis with the University of East Anglia and The Sainsbury Lab in UK, over his Strasbourg research lab with CNRS in France to his current ETH professorship in Switzerland. Voinnet even admitted to have manipulated data, for representative purposes. 8 papers were retracted with 18 corrected to date, and back in 2015 many protested that CNRS was actually too lenient. In fact, the then-CNRS president Alain Fuchs adjusted the rules on research integrity to avoid sacking Voinnet. The plant scientist was however virtually banned from returning to work in France and left stranded in his Swiss exile. His EMBO Gold Medal and his research funding was revoked (except that by ERC), and Voinnet became a persona non grata at conferences (read here and here).

This year’s second investigation by CNRS and ETH studied some (but not all) of additional evidence, but found Voinnet guilty of insufficient supervision only. This was supported by University of East Anglia, who announced that despite 2 retractions and several embarrassing corrections stemming from his PhD research, with exactly same photoshopped figures present in his PhD thesis, Voinnet is actually innocent (read here).

Following my prediction (or maybe rather my scheming conspiracies), Voinnet’s former right-hand man in Strasbourg, Patrice Dunoyer was found guilty of having faked data over the years and behind his master’s back. Dunoyer became the ultimate boo-man in France, Nature declared his former boss innocent and denounced the henchman’s conspiracy of forgeries. The publisher utterly forgot that they had no problem at all with accepting Dunyoer’s papers just a year ago. The traitor Dunoyer, carrier of CNRS Bronze Medal, was exiled from France to the other side of the world, to his home island of New Caledonia (as I originally reported). ETH made clear that no further misconduct evidence against Voinnet will be ever admissible, and the case is closed forever.

Voinnet himself got a lawyer to make CNRS lift the 2015 sanctions and started talking to press. He declared his newly discovered sympathy for PubPeer for its “help to clean up research literature” and modestly stated:

“I am not Dreyfus, I am not a saint, I just want the truth!”

He sure is no Dreyfus, even if some seem to think so. But Voinnet has all the potential to return to France and international plant science with a revenge spree à la Count of Monte Cristo.

Schneider, Enemy of the Republic?

France is now in a peculiar situation, research integrity-wise. And I admit it is mostly my fault, from my continuous Voinnet reporting on, which started almost 4 years ago, to the recent revelations of a manipulated figure in an old paper by French Minster for Research and Innovation, Frederique Vidal. I originally reported a duplication, which then proved a triplication.

The Ministry of Research reacted swiftly and announced to be suing Leonid Schneider, on national media in NewsTank. This is how the Government of France reacts to pointing out simple facts of a duplicated gel loading control:

“Charges have been laid against a scientific publication published 17 years ago and co-authored by seven authors, including the current Minister of Higher Education,
research and innovation. These unacceptable accusations are totally unfounded. The goal sought by their author is clear: to destabilize the minister. Faced with these accusations, the Minister is studying the possibility to institute legal proceedings.”

I send my two-fingered salute back.

Frederic Dardell, molecular biologist and President of ‘Université Paris Descartes where the work was done, then explained in the same NewsTank article and on Twitter, that even if Vidal’s gel image is triplicated it matters not. For one, it is a loading control, and those are anyway irrelevant. And also, the similarities are anyway accidental because using same samples on same gel apparatus produces similar-looking gel bands, if that gel apparatus is plugged in into the Infinite Improbability Drive (this outlandish theory was postulated before by the German biochemist Roland Lill). 

Dardell is also quoted in NewsTank with this:

“The university chaired by Frédéric Dardel has signed an agreement with Pubpeer because it hosts one of their developers. One of the two creators of Pubpeer – Brandon Stell – is also researcher in a research unit at the University of Paris.

For Frédéric Dardel, Pubpeer is an “interesting initiative” but the system “escapes
its designers” because “some, like Leonid Schneider, turn it into a court”.

CNRS, a fraudulent swamp?

Former CNRS president Anne Peyroche is about to be crucified while we still do not know who exactly faked the figures in papers from her lab. The CNRS head biologist Catherine Jessus is holding hands with everyone in power while 500 of her colleagues signed a Stalinist letter to abolish free press and to initiate purges for dissidents (read here). Among these signatories are the former CNRS president Alain Fuchs (who now calls for “presumption of innocence” instead of evidence of data manipulation), and Guido Kroemer, cancer researcher in Paris with CNRS background and an impressive PubPeer record which I recently wrote about. The current President of CNRS, Antoine Petit, described me and other PubPeer commenters as “assholes”, compared PubPeer evidence in Jessus and Peyroche affairs to paedophilia slander, while threatening to smoke out anonymous commenters using deep net surveillance (read here).

But back to Voinnet and the affair at the Institut de Biologie Moléculaire des Plantes (IBMP) in Strasbourg, where Dunoyer was allegedly busy faking Voinnet’s research data in total secrecy.

At the end of this post is a list of Voinnet corrections and retractions. But there are actually many more papers which were never corrected, despite serious PubPeer evidence. Like this beauty, Lecellier et al Science 2005. Dunoyer is last author, but let’s not hurry pointing fingers. Charles-Henri Lecellier was actually kicked out of IBMP by Voinnet himself for manipulating data in his lab. I am not sure what this proves these days, but anyway. Lecellier was sent to Montpellier, where he set up his own data manipulation laboratory in cancer research. When he was caught out by his collaborators in Luxembourg, he received highest protection from none other but Catherine Jessus (read the introduction part here). Lecellier never had to correct a single one of his papers, though he did feature as co-author on Voinnet’s first ever retraction, plus on another retraction (Vetter et al Oncogene 2010) which Jessus initially prevented, but which eventually happened due to my reporting.

If anyone should ever bother to investigate that Science 2005 paper (hint: not likely), rest assured that Lecellier will never be implicated in anything. He is well protected in French academia, while his less lucky data manipulating colleague Dunoyer had to resort to his impressive family connections to get a governmental job in New Caledonia.

There are more suspicious names. A Voinnet co-authored paper from IBMP, Dunoyer et al Plant Journal 2002 was corrected in 2015. The correction contains a fake figure, this what apparently also what the CNRS and ETH investigation from this year found out. It takes some guts and some very special sense of humour to fake a correction, and it is also impressive that the journal didn’t bother to check. Here it is, a section of background was mirrored and duplicated.

As the editor of the journal Lee Sweetlove explained, the correction was submitted by Dunoyer and the other co-first author, Sebastien Pfeffer, now group leader in Strasbourg, back then a PhD graduate at IBMP. 

This Dunoyer et al 2002 paper contains quite some data from Pfeffer’s PhD thesis from 2002. Including this figure, which somehow was reused in a different context in the paper.

As the highlighting shows, the photograph of the plant leaf is labelled differently in Pfeffer’s PhD thesis, it shows a plant treated with a different genetic vector  from what the same leaf in the Dunoyer et al 2002 paper stands for. Maybe an innocent mistake? But why was this other image from that thesis reused and not just relabelled, but also rotated 180°? Can this be an oversight also?

Maybe the journal will have to issue a correction to a correction now, and include this new material which one can hardly blame on Dunoyer.

Actually, Pfeffer had last year a retraction of his own, Pfeffer et al J Virology 2002, (see here for PubPeer evidence). But back when the retraction was steaming fresh and an evaluation travelling circus of Photoshop experts arrived to IBMP in March 2017, everyone apparently preferred to blame Pfeffer’s co-author on that retraction, the unlucky Dunoyer, and to keep the carrier of a running ERC Consolidator Grant out of the trouble. It might have counted that Sebastien’s dad Michel Pfeffer is an emeritus bigwig in Strasbourg and French academia, the senior even signed the Stalinist letter in support of research fraud.

Sans Dunoyer

Moving on with other examples of questionable co-authors on Voinnet’s papers. A retracted and very highly cited publication from his English PhD period, Voinnet et al Plant Journal 2003, has Susana Rivas as coauthor. That colleague from Spain, who used till recently to be CNRS group leader in Toulouse, France, was caught with much of fake data of her own, four more of her papers were retracted for data manipulation (see my reporting here). Her institutional website has been deleted now.

What about other retracted and corrected papers by Voinnet from his PhD period? There are no other suspicious names on it. Guess the conspiracy theorist will have to accept the responsibility of Voinnet here after all, because the only other common name is that of his PhD mentor David Baulcombe, who is now in Cambridge. Baulcombe allegedly used to joke that he “had to keep a close eye on Olivier from the beginning”, apparently not close enough.

Fun fact 1: the Voinnet affair initially started in late 2014 when some anonymous PubPeer sleuths started to screen Baulcombe’s papers.

Fun fact 2: Baulcombe was hosting Voinnet’s external investigator Witold Filipowicz right after he was done investigating at ETH(read here).

Fun fact 3: Baulcombe was acting as editor against PNAS guidelines on conflict of interest for the now retracted Voinnet paper Moissiard and Voinnet, PNAS 2006 (more here).

If we follow the conspiracy logic, whom to blame here? Dunoyer is not co-author, just like on several other retracted and corrected papers from Voinnet’s lab at IBMP and ETH.

Like this Ciaudo et al PLOS Genetics 2013, retracted for data manipulation after a failed correction: Dunoyer was not part of that either. Actually, why did ETH never even interview the first author Constance Ciaudo, whom they appointed as professor soon after she arrived to Zürich as Voinnet’s postdoc? Or another Dunoyer-free example, the horribly fake Deleris et al Science 2006, which even the benevolent ETH investigators set for retraction. Science‘s then-Editor Marcia McNutt explained she chose a correction as not to damage the careers of young scientists among the authors. Thanks to that, Angelique Deleris did rather well, she is tenured CNRS researcher and works under another Voinnet co-author who had to correct his paper in Science, Lionel Navarro. 

If ETH would not have permanently closed the Voinnet case now, we would also be expecting to see another paper corrected, or maybe worse: Mari-Ordonez et al Nature Genetics 2013., done in Voinnet’s ETH lab, also here sans Dunoyer. The first author Arturo Mari-Ordonez was just made group leader at Gregor-Mendel-Institut in Vienna, Austria, and also that institution said they will not investigate anything here. Maybe they would if one could somehow blame Dunoyer? Another problem is: Nature Genetics is a bit confused a journal, whose own editor seems to struggle with research integrity in his own papers (read here).

Then again, the paper Sansregret et al PLOS Pathogens 2013 was retracted after the Canadian first author Raphael Sansregret admitted to me in an email from September 2015 to have manipulated the Figure 1B when he couldn’t recover the original loading control. He also declared to me that that paper’s other problematic Figure 6 was manipulated by Voinnet and Dunoyer.

Shall we assume a mass conspiracy against Voinnet then? All his colleagues, lab members and collaborators unable to stomach his genius and his success, putting fake data in to sabotage his otherwise scientifically perfectly sound visionary ideas? If we go down that route, we must include the US plant scientist and whistleblower Vicki Vance into the circle of anti-Voinnet conspirators. Though it was a Dunoyer et al paper she kept rejecting as peer reviewer in 2004, it was Voinnet himself who kept resubmitting it while manipulating its data to adjust to reviewer criticisms. If we choose to believe that Voinnet is innocent, it must be obviously Vance who tried to sabotage his research here, no? That paper is now also retracted, it was in fact Voinnet’s first ever retraction.

Not a conspiracy, just massive corruption

Very likely not just Dunoyer but some others among Voinnet’s co-authors and colleagues engaged in data manipulations also. At IBMP, photoshopping seems to be a kind of strange culture, even the director Laurence Drouard is part of the problem here, if one studies the papers she published and theses she supervised. With such a stellar example as Voinnet to show the way to success, who can blame these academic members of French science elite? Well, we can, but very, very cautiously. They are powerful people.

Some others of Voinnet’s co-authors maybe just knew, and were happy to play along with his and Dunoyer’s creative figure preparation if it meant big papers in big journals and big academic careers for themselves. They are still guilty, even if they never faked a gel figure in Photoshop themselves. Knowingly submitting manipulated data for publication is still research misconduct, regardless of who actually did the faking. But the research community decided these academic careers are too important to dig all that up or to ask any questions. We saw the mind-bogglingly wrong reaction of French elites to the Jessus affair, where even a critic of Stalinists, Patrick Lemaire, spoke of “questionable methods that we all applied at this time“. Are French science elites unable to understand the concept of research fraud? Is Voinnet to be absolved of responsibility because everyone does it in France, and therefore he could never have learned the basics of research integrity?

This is of course nonsense, most French researchers sure know right from wrong, yet it is another question altogether which scientific method the French academic system rather rewards. The sad state of research integrity in Strasbourg, the Jessus data manipulation affair, and her own handling of Lecellier’s activities, offer some clues.

So now we are to scapegoat everything on Dunoyer alone, and get ready to celebrate the next Voinnet paper in a big journal.

I for one would rather hear out those who did not agree to Voinnet’s way of doing science. Where are those failed lab members of his? Probably not even in academia anymore.

---

Voinnet RETRACTIONS:

1. Brigneti G.et al.,EMBO J (1998)
2. Voinnet, O.et al., Plant J. 33: 949-56 (2003)
3. Dunoyer P.et al., Plant Cell 16: 1235-50 (2004)
4. Moissiard G. and Voinnet O., Nat.l Acad. Sci.USA 103 (51): 19593 (2006)
5. Dunoyer P.et al., EMBO J (2010)
6. Dunoyer P. et al., Science 328 (5980), 912 (2010)
7. Ciaudo C.et al., PLoS Genetics 9(11): e1003791 (2013)
8. Sansregret R.et al., PLoS Pathogens 1(9): e1005207 (2013)

 

Voinnet CORRECTIONS:

1. Ruiz MT et al, Plant Cell10: 937–946 (1998)
2. Voinnet O, et al., Nat.l Acad. Sci.USA 96:14147–14152; (1999)
3. Hamilton A.et al., EMBO J (2002)
4. Dunoyer P.et al., Plant J. 29: 555-67 (2002)
5. Parizotto E.A.et al., Genes & Development 18: 2237-42 (2004)                                                  2nd Correction 
6. Dunoyer P.et al., Nature Genetics 38: 258-63 (2006) 
7. Navarro L. et al., Science 312(5772): 436-439 (2006)
8. Deleris A. et al, Science 313(5783): 68-71 (2006)
9. Moissiard et al, RNA 13: 1268–1278 (2007)
10. Haas G. et al, EMBO J 27, 2102-2112 (2008)
11. Ciaudo C.et al., PLoS Genetics 5(8): e1000620 (2009) 
12. Azevedo et al., Genes & Development: 24 (9): 904 (2010)
13. Bennasser Y et al., Nat. Struct. Mol. Biol. 18, 323–327 (2011)
14. Jay F,  et al. (2011) PLoS Pathog 7(5): e1002035
15. Brodersen P.et al., Proc. Nat.l Acad. Sci. USA 109: 1778-83 (2012) 
16. Gibbings D. et al, Nature Cell Biology 14, 1314–1321 (2012)
17. Schott G. et al, EMBO J 31, 2553-2565 (2012) 
18. Boccara M.,et al., PLoS Pathogens 10(1): e1003883 (2014)

---

Update 9.01.2018. News Tank published an interview with Olivier Voinnet, who is portrayed as the most honest researcher ever, a victim of Dunoyer’s fraud and Jessus’s scheming, and an expert in research integrity. The interview is heavily paywalled, not even searchable on Google, but some kind person uploaded a copy here. Please do not download it it will make News Tank sad, they value their copyright.

Update 18.01.2019. Surprise! Well, not really. Turned out same Chicken of Dishonour Legion, Francis-Andre Wollman, also acted as original Voinnet investigator in 2015. This was revealed by Le Monde,

Basically, Wollman found Voinnet guilty of fraud for very similar data manipulations found later in Jessus’ papers. Yet there, Wollman saw nothing untoward whatsoever. Same kind of gel band copy-pasting can be either fraudulent or perfectly scientific, depending on who did that.

I think the best way to celebrate this, is to flag a gel duplication in a paper Wollman co-authored last year, Wittkopp et al Plant J 2018

Update 23.01.2019. Another Wollman co-production, Cardol et al PNAS 2009.

---