Regarding the Olivier Voinnet scandal and a recent data integrity case in Germany (marginally featuring a current group leader of the John Innes Centre in Norwich, UK), I had an email exchange with the director of this British plant science research centre, Dale Sanders. It started with Sanders demanding of me to cease-and desist from ever associating the misconduct-tainted cheater professor Voinnet with his John Innes Centre (despite that connection being very well-documented), and ended with Sanders calling me an “internet troll” and decreeing that my Voinnet reporting is not worth ever being read because it has not been peer reviewed.
The whole case revolves around the former star of plant sciences and current ETH professor Voinnet, though it is actually not at all about him or his own data manipulations. In July 2017, I brought an article, where concerns about the data integrity in publications of a very senior and influential German yeast biologist, Roland Lill, were raised on PubPeer. A past member of Lill’s lab, his PhD student Heike Lange, is now tenured researcher at the same CNRS institute in Strasbourg where Voinnet did most of his data manipulations. Lange has a number of papers which contain what clearly looks like duplicated western blot bands, she and LIll went to PubPeer to declare that the bands were never duplicated (details in my article). Yet Lange and Lill never substantiated their claims or showed any original gel scans for their common papers, which lets one wonder if those actually exist.
It turned out somewhat differently with another former PhD student of Lill, Janneke Balk, who is now the above mentioned group leader at Sanders’ John Innes Centre. Two common papers of Balk and Lill were flagged on PubPeer: Balk et al EMBO J 2004 and Balk et al, Mol Cell Biol. 2005. She has not managed to address the concerns about the former yet (more about it later), but she did reply on PubPeer to the evidence of gel band duplication by admitting the copy-paste, posting the original gels and explaining at length exactly which band was duplicated and why:
The action by Balk to provide original gels and explain her scientific rationale in detail is indeed very laudable, and her institute director Sanders instructed me in the email below that this closes the issue and ends all concerns. In my view however, even if there is (currently) indeed no reason to suspect in this particular case of Figure 6C anything but some occasionally very poor scientific methodology and absent research integrity training in the Lill lab, the debate should not be over yet. In fact, Balk admits to have misled the editors, reviewers and the scientific community by having duplicated gel bands, spliced them in Photoshop in a way to avoid detection, by having failed to use the correct genetic wildtype controls, and also, among other issues, by pretending to show what looked like one equally loaded gel where there were actually two separate gels without any loading controls. I formulated these points in my own Pubpeer reply.
The original PubPeer post by Balk does explain why she manipulated data in the way she did, but this information must also be made available to all readers of her Balk et al 2005 paper, so they can form their own opinion on its appropriateness. Yet Balk and Lill do not mention anything about the need to contact the journal’s editor or that any correction to their paper might be in order.
This was why I contacted Sanders’ John Innes Centre on August 12th, asking them to comment on Balk’s issues showcased on PubPeer. I brought into their attention the past case of Voinnet, who did his PhD under David Baulcombe at the associated The Sainsbury Laboratory (TSL) in Norwich, and how his PhD thesis, despite its admitted data manipulations, was never really investigated by the University of East Anglia, which is the academic partner of the John Innes Centre and TSL. This was how Sanders replied on September 9th:
Dr Balk has in my view handled this issue in an exemplary fashion.
Concerning your message, there are two surprising errors, given that you are an investigative journalist:
- the Voinnet affair at JIC: Voinnet was never at JIC; he was at The Sainsbury Laboratory when in Norwich.
- how his PhD thesis containing such data manipulations was never investigated: the PhD thesis was comprehensively investigated, and my belief is that should you wish to contact the degree-awarding body (University of East Anglia) you can obtain a statement on the outcomes of that investigation.
It would be good to receive your assurance that you will not perpetrate these erroneous comments in the future.
It sounds somewhat like a cease-and-desist letter, and Sanders never denied this suspicion of mine. I informed the John Innes Centre director that there still are open questions regarding Balk’s Figure 6C in Balk et al MCB 2004 , and reminded him of her other flagged paper, Balk et al EMBO J 2004 (PubPeer evidence here):
But in the end, Balk data integrity concerns were never the real topic. Sanders went, as his first email made clear, straight into the Voinnet affair. I pointed out to him that as for his point 1, he would have to take up issue with Prof Voinnet himself, who writes on his CV on his own ResearchGate profile:
Dec 1997– Dec 2002 John Innes Centre · The Sainsbury Laboratory · United Kingdom · Norwich
Moreover, it seems John Innes Centre used to expect back then (or at least did not mind) that all papers from TSL at that time featured it as additional affiliation. This certainly helps when central funding is applied for and list of publications from your institution needs to be provided. In fact, all Voinnet papers from Baulcombe’s Norwich lab declare, like this one: “Sainsbury Laboratory, John Innes Centre, Colney, Norwich NR4 7UH, United Kingdom”. Wikipedia states: TSL “was constructed at the John Innes Centre” in 1989 and “participates fully in the life of the John Innes Centre and the University of East Anglia”. But obviously, other past TSL researchers are meant to be included, not Voinnet.
Sanders hence wants both his cake and eat it. He is perfectly happy with “good” TSL papers wearing their John Innes Centre affiliation, but he demands of a journalist to promise never to mention problematic cases like Voinnet as ever associated with his Centre, despite all the evidence to otherwise. And as fro point 2, I pointed out to Sanders my two earlier reports (this and this) where I submitted three Freedom of Information Inquiries to the University of East Anglia about the outcome of a possible Voinnet thesis investigation, and was told to get lost. The university even refused to say if there ever was an investigation. An investigation which existence was denied and which never annoucned any results or had visible consequences (Voinnet’s doctorate is perfectly fine, thank you very much) simply did not happen. Sanders however, who together with his John Innes Cenre has nothing to do with Voinnet at all, seems however to know very well about the thesis investigation of the latter’s PhD thesis, which existence and successful conclusion the director now officially confirmed. This was what Sanders wrote to me next:
Your response was not one worthy of an investigative journalist, as you style yourself.
An investigative journalist would check facts independently and not rely on web links for their sole source of information. Internet trolls do the latter to build one-sided cases.
Here’s a simple suggestion regarding my second point: contact the University of East Anglia directly to establish whether they have carried out an investigation on Voinnet. You should receive a simple Y/N answer that should prove you or me right. Then please let me know the outcome.
Having said this, I do think PubPeer has an important role. I makes scientists think critically about responsibility in presenting their data. Shrill and inaccurate comments [as according to sanders this one is, -LS] devalue PubPeer’s role.
I pointed Sanders again to my published previous experience of trying to get any information from University of East Anglia regarding Voinnet’s thesis investigation, and regretted he could not find time to read my relevant articles I sent him (this and this). I also announced that I intend to publish his emails to me. This was how Sanders then ended the email exchange:
“That is absolutely fine!
No, I don’t have the time to read your “articles”. Who peer-reviews them?
I hope you also post this email that reflects on our correspondence to think that you have all the characteristics of an internet troll as defined by Wikipedia
You seem to display every symptom.
You are clearly not interested in science or its ethics.
The conversation is therefore over.
Actually, my journalism has been peer reviewed with a scholarly journal once, and failed spectacularly (regarding trachea transplants, story here). Strange as his accusations of trolling are, Sanders is however not alone in his views that my Voinnet reporting over the years was nothing but fake news of an internet troll. His French counterpart, Catherine Jessus, allegedly said that my reporting about Voinnet’s past Strasbourg institute IBMP (here) and its director Laurence Maréchal-Drouard (here), was “unreliable” and not worth acting upon. Which probably means either that my (or in fact PubPeer-posted) evidence of data manipulations is unreliable and those issues do not exist, or that I somehow smuggled those manipulations into the IBMP papers of Drouard and others myself. Meanwhile, Jessus own papers face intense scrutiny (here), by readers of my not-peer-reviewed troll site.
Sanders for his part made perfectly clear how he will respond to anyone trying to mention any research integrity concerns in connection with his own John Innes Centre. Don’t mess with Sanders.
Update 15.09.2017. A reader forwarded me this evidence of inadvertent image duplication in a publication by Sanders lab, which is listed 2nd among selected publications on its website:
Update 6.10.2017. The University of East Anglia denied for the 4th time my FOI request, and refused to share any information about the existence of the investigation of Voinnet’s thesis. Here is the last refusal, and now Sander’s cease-and-desist demand I admit that Voinnet’s “PhD thesis was comprehensively investigated” is the only proof that investigation ever took place. However, given its secretive nature, we can either trust the “authorities” that there was indeed nothing to see there, or tell them to shove their secrets, and form our own opinions. Here is the evidence.
If you are interested to support my work, you can leave here a small tip of $5. Or several of small tips, just increase the amount as you like (2x=€10; 5x=€25). Your generous patronage of my journalism will be most appreciated!
3 stars for Full Professor and ribbon bars for publications !!!
Directors of huge centers must have 5 stars
LikeLiked by 1 person
More from Weizmann Institute – note that a new study just published in Immunity ( http://www.cell.com/immunity/fulltext/S1074-7613(17)30318-7
) debunks a series of high profile papers in Nature and Cell (at least 3) from Eran Elinav – Weizmann institute who did not bother to use litter mate WT controls:
Pingback: Christmas messages from Professor Turner, his ex-protégé under investigation Tiwari, and Elsevier – For Better Science
Pingback: Olivier Voinnet: not guilty in past, present and future – For Better Science