We have an update on the case of Pankaj Dhonukshe, a scientist about whom we reported in November. Utrecht University has found that Dhonukshe, a former researcher at the Dutch university, committed “a violation of academic integrity” in work that led to a number of papers, including one published in Nature and once since retracted from Cell.
Here’s the university’s statement:
Based on advice issued by the Committee for Academic Integrity, the National Board for Research Integrity and a supplementary investigation by a technical committee, Utrecht University has decided that a violation of academic integrity has been committed by cell biologist Dr P. Dhonukshe. This finding does not apply to the co-authors of the articles that were the subject of this investigation.
In the figures of four articles of which Dr Dhonukshe was the first or last author manipulation in some form was discovered. The irregularities include the fabrication of data that were presented as actually obtained research findings. Already during the course of the investigation the researchers were requested to withdraw two articles from publication, as the articles obviously do not satisfy the university’s academic standard. One of the articles has already been withdrawn. The university published an announcement (in Dutch only) on this matter in November 2013.
Investigative process
Dr Dhonukshe worked at Utrecht University until January 2013. In April 2013, the university received a complaint from two Canadian scientists concerning possible irregularities in an article on cell biology research published in the journal Nature. Following this, the complaint was investigated by the university’s the Committee for Academic Integrity. In response to this, the Executive Board formed its initial judgement in September 2013. This in turn prompted some of the parties involved to ask LOWI for a second opinion. At the same time as the initial judgement was reached, a technical committee was tasked by the university to investigate other articles by Dr Dhonukshe, in order to assess the extent to which these too could be found to contain irregularities. In February 2014, at the request of the Executive Board, the Committee for Academic Integrity issued supplementary advice based on the report of the technical committee. LOWI also issued its advice in February 2014. Based on all these data, the Executive Board was able to reach its final judgement.
DUB has more details (in Dutch).
Dhonukshe, with whom VIB Ghent parted ways last year, sent us the following, which he had sent to Utrecht yesterday. He also sent us a February 3 letter to Utrecht, which can be read here.
I strongly oppose to the final decision of Utrecht University.
I strongly urge you to remove the word ‘fabrication’ in the publication documents and reconsider the decision.
Fabrication?
This decision is referring to ‘fabrication’.
As per the rules of Utrecht University ‘Klachtenregeling WI 1 september 2012_EN def” on page 6 fabrication is refereed as:
– fabrication: introducing fabricated data
The fabrication or invention of data which are presented as the actual findings of research. This goes to the heart of science: the process of establishing truth.Further, you are referring to the principle of reliability ‘The Netherlands Code of Conduct for Scientific Practice’ (principle II). As the word fabrication does not exist in the Code, the decision is groundless.
Erroneous duplication of image is no fabrication by any means. I did not draw (fabricate or invent) the images as one drawing a painting. They are factual data.
The statement of ‘fabrication of data that were presented as actually obtained research findings’ is baseless. This does not exist in any of the documents generated during the whole investigation process.
Speculation
As UU says ‘Based on its investigation, the committee formed the opinion that due to the nature and frequency of the identified manipulations and to the fact that these occurred throughout a number of successive years – manipulations which for that matter were not discovered in the articles of which Dr Dhonukshe was not the first or last author – these must have been at least in part deliberate.’ This shows that UU decision is speculative without a direct proof of fabrication.
Therefore, I ask the Utrecht University not to publish the decision and take the facts that I describe in the letter from 3th February sent to Utrecht (Mr de Bok), see attachment, into a reconsideration of the decision.
New key facts not considered
It is a specific failure that this letter has not been considered in the decision of Utrecht University. These new facts are key in the fundament of the decision and therefore require a reconsideration of the decision.
This letter addresses all the points raised by the Technical committee while I did not get any feedback on this. This letter has been completely neglected by the Utrecht University. My letter addresses all three points raised by the technical commitee:
1. gel band issues
2. microscopy image issue
3. duplication issue.
Further, Nature 2013 and Cell 2012 have confirmed irregularities but not the other papers. The conclusion of the Technical Committee that “the identified manipulations and to the fact that these occurred throughout a number of successive years, these must have been at least in part deliberate.” is wrong. My letter shows that many of the irregularities are not from my data and have been made by other researchers.
Procedural mistakes
Further, there has been number of procedural mistakes made during UU procedures without an apparent transparency that I questioned all along.
The complaints refer to mistakes in two papers. This may never be a ground for accusations of manipulation when mistakes are at stake.
Research and presentation
My vision of science was that truthful findings (content) that bring us to a next step of our knowledge are the most important. I did not perceive the details of presentation of the findings to be as critical as that of the actual findings. This resulted into the presentation mistakes that no one brought to my attention before the mistakes were brought up. Based on the last one year’s process I have learned that presentation is also important in addition to content. I shall incorporate that from now on in my perception of science and presentation of my findings.
Maybe I am stating the obvious here but this defence seems utterly disingenuous. Ethical issues around data presentation have been prominently discussed, and not least in these higher level journals, which have all had explicit policies on image handling for several years now. Ignorance or naivete is absolutely no excuse for a scientist of Mr Dhonushke’s standing to have engaged in this practice. I will be interested to learn how many other of his papers feature similar liberties with “presentation”. Does anyone know what is happening with his ERC grant, which was presumably significantly based on the two papers in question?
More worrying is his insinuation that manipulations were common practice in the labs in which he worked. “Fig. 3D of Cruz-Ramirez et al. Cell 2012….These data are also not from my work”.
Insinuation is such an understatement… As a defensive move Pankaj decided to turn the s*** ventilator on.
The letter clearly states that figure manipulation is a common practice, that manipulated figures were used in several published papers, and that misrepresentation of data shouldn’t be considered a big deal because the message is somehow more important than the actual facts and figures used to support the claims…
The letter shows no acceptance of fault play and a clear message to his former colleagues. Do not mess with me cause everybody cheats…sometimes.
Pure waste of ERC money
Perhaps the culture of the labs in which he worked encouraged the idea that ‘the message is somehow more important than the data’. Certainly his collaborating labs have published some stinkers recently. Hopefully this episode will improve scientific rigor by encouraging the plant community to contribute to online resources such as BioRxiv and PubPeer.
I am convinced Nature would be happy to publish a Corrigendum, where he will be allowed to make new figures to replace the ones in question, as long as he can pretend that he cannot find the raw data anymore…
An interesting example can be seen here in the recent Nature, where several copy/paste and duplication issues were fixed in a Corrigendum: http://www.nature.com/nature/journal/v506/n7487/full/nature13003.html
I still believe that if authors cannot even prepare their figures properly, how can I trust how they prepare their experiments?
I find it difficult to decide what we’re looking at here: a poor researcher casting the blame on his coauthors or an average researcher correctly identifying himself as a scapegoat, and who is given the blame for categorical failures in his research group.
I think the only way we can do more than guess is to study the reports prepared by the various institutions, although I fully expect the university will keep its reports secret. LOWI should produce something, though.
The fact that the papers in question span PD’s time in at least two different groups and his time as an independent group leader suggest the former over the latter.
Indeed, if you are going to discredit someone it is far more effective if you consider the body of their work, not just the recent pieces.
I agree.
LOWI should publish a report. and not some wordy bureaucratic 100 page report, but a concise one with methods, results and conclusions, so that we can look at the evidence and form our own opinion.
“We looked into and its all his fault, just trust us” is not enough!
I do not know how putting dirt on others can make oneself clean. OK, there are other people doing similar wrongs, but you are also one of them. Plus all this business of not knowing the rules and unintentional ‘mistakes’. If after a PhD, 2 post docs and several years as an ‘independent researcher’ you do not know how to read ‘ instructions for authors’ then there is something fishy about you. The are problems with your work : period. To know if they are intentional or not one has to read your mind, which is not possible. Moreover naivety is the old sheepskin under which manipulators always masquerade themselves.
Don’t be mislead by these excuses. Picking the right image, counting right roots. Many people have seen this ‘style of work’, where publishing paper was above all: quality, integrity, ethics. Of course he is not the only one to blame, but does it mean he is clean?
Certainly he is not clean, but big questions remain about the labs involved.
Why did only two co-authors realize there is a problem?
Why are the conclusions from the 2008 Nature paper okay, despite no original data being made available? (From the dutch website: “On the articles from 2008 and 2010, you can still perform debate, because there are far fewer errors. Manipulation also not affect the conclusions of those articles.”). This makes no sense at all.
How do the labs involved respond to the allegation of manipulations in papers and figures that PD did not author?
It would be interesting to hear what they have to say. However, given the general tone of his defence it is my impression that the claims that certain data were not “from his work” also probably need a closer look. While these statements may be technically true, it is also possible that as a more senior lab member, a certain degree of “mentoring” in image selection and processing was provided.
It also makes me wonder – clearly not clean but there could be some degree of truth in his protests that certain subjective practices are widespread. Could the actual problem that he just didn’t know how to play the game well enough to stay on the right side of the line that separates undetectable bias in image/sample selection (no doubt common in these types of papers) from actual image manipulation? Reflecting a combination of naivete, stupidity and arrogance?
He went from the auxin field into another scientific area (microtubule biology). It was researchers in microtubule biology that expressed concern and started the investigation. Make your own conclusions about what is acceptable practice in the auxin field. Certainly, as ‘Sean’ pointed out in another thread there is a lot of subjectivity in confocal imaging of cells.
But it is tough, as the glamour magazines want a complete story. There must be a temptation to fill the gaps of your data to make the complete story. There are limited jobs available. To get a job you need a glamour publication, and the good graces of the existing researchers in the field.
just curious to know. The person who is involved here has worked in different places. Potentially, he had many co-workers some of whom are his co-authors. When you are working together in the same laboratory, you spend time together – going for coffee/tea, dinner or hanging out somewhere. These co-authors (who are working under the same supervisor) might definitely aware of things going one. Why don’t they speak up here (anonymously). Some co-authors might have benefited heavily from high profile publications right? Are they in trouble or not? when you have couple of co-author papers in high profile journals your credentials also go up. When those papers are retracted – does your status get affected? Any evidence of collateral damage?
Just for your information. Pankaj Dhonukshe is clearly coming from microtubules to auxin in his career. So it is completely other way round. But more importantly, I am always worried when people blindly shooting on auxin field. It is unfair and very dehonesting. There is nothing like acceptable practice for one field and inacceptable for the other! We all are on one board, we all should do our best to be unbiased when doing reasearch and reporting about it.
As a science journalist for the Dutch biweekly for biologist I have written an article about this case for our March 1 issue (in Dutch, I am afraid). One thing that disturbes me is that we have to rely on press releases of the university (‘yes, he did manipulate in 4 articles’) and of the culprit (‘sorry, I did make mistakes, but not with bad intentions’), while we can not do our work as press properly. We can not check facts in the underlying reports, because they remain ‘confidential’. The LOWI-report will be made public after six weeks but anomynized (and summarized?) which makes them hard (=almost impossible) to read. Stating a hard verdict as misconduct and not presenting the facts that sustains is a strange combination: especially in science. It feels like a verdict without a real explanation. This calls for an more transparent way to deal with such cases: maybe a trail section in PubPeer? We should allow more oxygen to get in, as part of the normal scientific procedure, to prevent scientist being tempted and ending up with rotten results.
The 2008 Nature paper is “Generation of cell polarity in plants links endocytosis, auxin distribution and cell fate decisions.” It contains egregious fabrications in figure 1a and S2a. The technique is FRAP, which in this case stands for Fluorescence Recovery After PhotoShop. Briefly, the 2 minute post bleach is simply a black box. And the remaining time-points are actually just different planes from the same confocal stack as t = 0. Yes, it’s that flagrant. So he took one stack, never actually even bleached anything, went to PhotoShop, and performed “FRAP.” LOL.
I am sure he’ll say: ‘ sorry, I did make mistakes, but not with bad intentions’ or ‘not my data’ or ‘ I was not aware of the procedures’ or ‘It is only .005% of the data and does not change the main findings of the story’. LOL indeed.
as i said earlier, co-authors should have voiced their concerns then and now..when the papers are being reviewed and under revision with Nature and Cell – everyone gets mileage isn’t it…..
They missed it. The entire auxin field missed it. So your logic is that once the crime is committed, the person is no longer responsible if they get away with it. Do you really think that the coauthors saw this fabrication, but decided to hold their tongue because they so desperately wanted a nature paper? And the reviewers. What reviewer on this planet would pass up the opportunity to zing a cheater?
no. I did not say that the person is not responsible for it. Can you guess the collateral damage? When he says that he did not do it for some papers – then who? You don’t get my point! With success of publishing in Cell and Nature, all authors celebrate right – everyone gets mileage! When it is undone – imagine the damage to their career as well!! Well – check the lab’s track record and talk about the reviewers passing the paper….it is not a small laboratory and PI is topnotch scientist.
Dear KK,
The problem is, most of the time a paper is send to collaborators (co-authors) they mostly check if their own data is represented OK in the final draft . I guess the rest of the paper is left to good faith and it is believed that your collaborator is truthful ( the faith is stronger if your contribution supports the lead author’s story). In ideal case, the collaborators should do independent repeats, but in this rat race (as JW has pointed out in a previous post) who has the time for that? Also, in this particular case, the manipulations are very very subtle.
On a whole what do you think? Many people,including Pankaj Dhonukshe and the famous Japanese researcher Shigeaki Kato (retracted total of 43 papers) claim that the manipulations were only in minor parts of the data and because the whole paper is retracted even some imp key findings are also trashed. Should these people be given a chance to redeem themselves and publish the true findings as a corrigendum.