We’ve learned more about the circumstances behind a Cell retraction that we covered last week.
First, one of the two corresponding authors left the institution where he most recently worked. Belgium’s VIB Ghent told us that Pankaj Dhonukshe was no longer employed there and said:
Any work performed during his very short stay at our institute, did not lead to scientific publications.
We parted ways this summer as VIB decided to stop the employment agreement with Mr. Dhonukshe, based on the VIB standards concerning scientific quality, integrity and leadership. The work described in the Cell paper that has been retracted was not performed at VIB.
Dhonukshe declined to comment for this post, citing the ongoing investigation at Utrecht University, where he was working when the research was completed.
Second, we learned that the retraction notice initially described the problems as “manipulations,” while the final wording is “mistakes.” Ben Scheres, the other corresponding author, tells us:
Indeed in the retraction notice draft that was considered appropriate for signature by the majority of the authors, including me, we used the word ‘manipulations’. You should contact Cell to find out how it got changed. Proofs of the retraction notice were sent to the corresponding authors, and I noted on hindsight that the change was in the proofs; it was not explicitly flagged and I missed it. When the change was pointed out to me by coauthors after appearance of the retraction, I immediately contacted Cell to question this change and cc’d the concerned coauthors. This inquiry to Cell is still pending.
Cell editor and Cell Press CEO Emilie Marcus tells us:
The preliminary findings of the university investigating committee are confidential and therefore cannot be quoted here, however as communicated to Cell by Utrecht University officials their investigation was unable to identify intentional manipulation for the data presented in this paper; the findings of the investigation are currently under appeal. An earlier version of the Retraction text did contain a reference to “manipulation” but was revised in keeping with editorial policy to be more closely aligned with the findings of the institutional investigation. The final text was sent to both corresponding authors (PD and BS) prior to publication and no further corrections or changes were requested.
We acknowledge that there are significant controversies and concerns surrounding one of the authors of this paper and the Retraction statement of this paper alone will not address all of these issues appropriately. That said, we feel the current Retraction statement accurately reflects the information available about this specific paper and states clearly that the “original data were processed inappropriately.”
Please see an update on this post, with results of the Utrecht investigation.
Corrections, mega corrections, mistakes, errors in assembling figures. Slowly but surely, the connivance of the scientific publishing institutions in the not maintaining standards comes out. I can only wonder if it was a new hiring, lacking cynicism and full of enthusiasm who called the spade the spade and used the work “manipulation” only to be overruled by someone higher up?
i notice that other member from the same laboratory Kalika Prasad who is the 7th author in this paper has landed in a good institute and good prospects for him/her.
Is there reason to believe that this other person may have been complicit in some way?
There are 18 authors listed. Is it fair to post a comment about one in particular without any clarifying information?
I also hasten to add that I have no vested interest. I do not know this paper, the authors or the field. My response is based purely on what I read here.
My mistake. I thought it is the 7th author they are talking about in the previous blog. http://retractionwatch.com/2013/11/22/authors-retract-cell-paper-amid-ongoing-investigation/
Someone mentioned about 1st, 2nd and 6th authors..i retract my comment.
Look the RT PCR in that paper! Play with the contrast in image J. You’ll see that the space surrounding a given band is completely black. Meaning no background from ethidium or sybr or whatever DNA stain used. That is impossible to achieve. There is always some background. I’ve been doing this stuff for about 15 years. Then the editor of Cell says “manipulations” aren’t consistent with the investigation. Hmm… the only way to achieve those results is… wait for it…. manipulating pixels. Scientific publishing is big money and a house of cards. It looks better for Cell if they come out and detect “mistakes” instead of having their review process exposed for the sham that it actually is.
“…instead of having their review process exposed for the sham that it actually is.”
Amen. I’m nowhere near your level of experience (just finishing a PhD) but, in light of the fact that this whole show is bankrolled by the taxpayer, the amount of disgusting behavior of all kinds in this field of work is just absolutely repelling.
Not just scientific publishing is a house of cards …
That was a narrow escape for Ghent. Everything staying safely over the border in the Netherlands.
Doesn’t VIB reward its researchers who regularly publish in Cell and Nature? Top management must be disappointed as, at first glance, this would have been exactly what they were looking for with four Cell and Nature paper co-authorships in three years. What a let down for the running dogs of citationalism!
If you publish 4 excellent papers in 3 years there is a question to ask: what about data quality? Research need time and when you are in hurry… well we all see the result! I hope the truth will come up and there will be no mercy for this kind of actions.
Another gem is the fact that contrary to Ghent’s statement, in fact he did publish while at Ghent: http://www.ncbi.nlm.nih.gov/pubmed/23903948
This whole thing disgusts me. I can’t wait to see how Nature spins their likely retraction of the Kakar et al 2013 paper, which appears to contain such brazen fabrications that it makes you wonder what drugs the reviewers were on when they reviewed it. And then the 2008 Nature paper…PD fooled not only his co-authors and reviewers, but has successfully deceived the entire field of plant biology for over 5 years. (hint: look at figure 1a and s2a, which show “FRAP”. Note the post bleach is just a black box, and 100 and 200/300 minutes are just different image planes from the same cell at the same timepoint “zero”. For shame.) It seems the only critical thinkers left in the field are those we get for reviewers!
Very informative. Although to be fair to the reviewers, they are probably PIs who haven’t done much of the new types of experiment. They take the claims of the papers at face value. No one has time to carefully analyze each panel in figures. Recent Cell and Nature papers have umpteen panels to each figure (the Cell paper has two figures with panels from A-Z – they ran out of letters in the alphabet…) Add the mass number of supplemental figures and you get a complicated mess.
From what you say there is doubt on previous first author papers from Pankaj Dhonukshe. Perhaps even doubt on all the papers on which he is a co-author. What was his contribution to each paper? It is hard to tell the contribution of a single person when there are so many co-authors.
Totally agree. The reviewers don’t have time to properly analyze data since they are probably PIs, and haven’t been in lab for years. Same thing on the “supplemental” figure thing. It’s getting way out of hand. A lot of times in fact, the supplemental data is actually an independent point that should be in the main text, but because the journal is Nature or Science and only allows like 1 square inch for figures, they just stuff it in the supplementals. I feel it is part psychological in the sense that it not only looks impressive to have lots of supplementary data, but it also provides distraction and camouflage for any…ahem…”mistakes.” Same thing with stuffing 20-30 authors onto these papers; looks impressive; you get an immediate unconscious impression when you see the paper that it must be a big deal. As far as author contributions on the 2 Nature papers, I don’t know, but the common denominator is: Dhonukshe.
May be you are true about Dhonukshe. But about the 2008 nature paper,how could you be so sure that 100 and 200/300 mins post-bleach are just different image planes from pre-bleach. I am not defending Dhonuskhe, but simply asking from academic interest if there is any tool to detect such a thing.
Out of curiosity, I just had a look at Dhonukshe’s 2008 nature paper. I was surprised this was not discovered earlier by people. It is easy to see there are two black boxes within cells in figure 1c (120min and 180min).
No tool at all to detect this kind of thing, unfortunately. The only reason I was able to spot it is because I have been imaging these same things for years. By rolling to different planes within a cell, the degree of apparent PIN polarity changes, appearing most polar at the surface plane since there are no shared sidewalls with neighboring cells in contact. At the median plane, the signal is always less polarized, often even appearing all around the cell, as he shows. What is funny about this experiment is that even if he had performed it, the logic is flawed. In concept, it is a pulse-chase kind of experiment, but PINGFP is CONTINUOUSLY expressing, so it’s not going to first go to all cell faces, and then disappear from the sides! Photoswitchable FPs would be the only way to do this experiment with any meaning. But of course, Dhonukshe never did the experiment anyway. Layers of comedy here.
Ha ha ha… whistle blower, you are absolutely spot on. PIN2 is continuously expressing and being endocytosed. So,the idea of initial non polarity and then achieving polarity by endocytosis makes no sense at all. The expt is flawed by logic and getting this kind of result it clearly shows that that it is manipulated. Apart from using photoconvsersion, one can use inducible promoter also work, which was clearly not the case here (pin1:pin1-gfp). I wonder what the reviewers are thinking.
Couple of typos in my reply, sorry. I did not mean PIN2 in the 2nd sentence but any PIN.
I don’t know about comedy, more like tragedy. There seems to be a tendency to make a pretty story or model about how biology works, and then cherry pick data to fit your story. I think this describes the retracted Cell paper and some other recent high profile plant papers. I doubt the co-authors were skeptical of data that reinforced their story. Perhaps PD doesn’t think he did anything wrong, still believing the model is right and the data just needed help.
Is it just PD? A strange moment happened at the Auxin meeting last year when a speaker directly challenged the experiments from a 2010 Cell paper (10.1016/j.cell.2010.09.027). PD is a middle author on this 2010 paper, but it is the same lab as his 2008 Nature paper. I thought this was just bad blood between the two researchers as they previously worked together, but perhaps not. What did you think of that paper? You seem to know the field.
I received an email from a website concerning a 2013 paper from the same group.
https://pubpeer.com/publications/B5B747042EF3AA0864621F6033F6F1
1 strange looking plant out of 184 mutant seedlings is a ‘representative’ phenotype? I guess the 183 wild-type looking seedlings were not sufficient to disprove their hypothesis on this gene’s function. It does make you wonder how ‘representative’ are the confocal images in their papers? Is 1 image out of 184 enough to be ‘representative’ if it supports their story?
Dear JW,
Many thanks for the update. About the 2010 Cell paper: as a classic Cell paper is, it’s a jumble of connected and unconnected facts (quite easy to confuse the reader). I guess the part that auxin inhibits endocytosis may be true as it is also supported by other independent groups(http://www.plantcell.org/content/21/2/568.full). But they directly contradict on the role of SCF/TIR signalling. Who knows who is right?
There are some common concerns I have though relating to such expts:
1. PIN polarity as mentioned above could be manipulated by imaging different planes. The reviewers should ask for a Z stack.
2.Use of drugs in excess. Some papers of the group use Tyr-A23 at 30um, some at 45 um some at 75um and so on. I mean if the minimum works then stick to it. Also use of Hub, DN-Ara7 etc. Although specific than chemicals, they after all kill the plants. It’s very easy to image sick plants and show any cell biological event is hampered. They show it with endocytosis, I can show it for glycolysis.
The field is saturated with studies on PIN polarity, everything under the sun seems to block clathrin and hamper pin polarity. Even the figures from different papers are following the same pattern. I mean ……..move on !!
AMEN!!!
I talked to my advisor about this an extremely long time ago, and several other similar papers. He told me it was human nature and that the reviewers are biased. I called it nepotism, and I think it is rampant in science. I am almost completely over all of this and am beginning to lose interest in science in general. the gatekeepers don’t even read the papers. I can not reliably make citations for my own research project and subsequent publications under the circumstances. I thought I was losing it and I was the only one that noticed this, but after seeing my perspective echoed here I really don’t know what to believe. Why do I even bother to be a part of the scientific community if the highest impact journals have standards like this. What is the point? I might as well just go make money elsewhere in industries I understand instead of passing through these gates of nepotism.
Take a look at this: http://link.springer.com/article/10.1007%2Fs10529-010-0290-0
I’ve detected at least 100 errors, some really major, in this paper. An anonymous report was made. The editor stated that old papers cannot be questioned, that anonymous whistle-blowers were harassers and that, despite all the evidence, that he took full responsibility. Silence masks the unprofessional and most likely biased peer review that took place. Why can editors and publishers investigate authors but why are authors and the scientific community not allowed to investigate the editors, request copies of “peer reports” and examine the publisher in detail?
Science publishing is slowly starting to expose itself to be a big fat business sham, and science, even good or honest science, is being dragged through the mud. NO longer is it becoming possible to distinguish correct from incorrect science, honest from dishonest scientist, reputable from fake journal/publsiher.
HM Xiao got the emotion spot on. Be in science, for what?
Sadly…this is true. Quite often the name/surname of the corresponding author decide the acceptance rather than the data. One solution is to introduce anonymous review on both sides (the reviewers won’t know the authors name), but I think neither the journals nor all of the authors will agree to that.
Double blind reviews are a nice idea but would never work in practice. The paper would be anonymous in name only, as determining who did the research is easy.
Most people receiving negative reviews are convinced they can tell who wrote that ‘anonymous’ review.
My problem with open reviews is the ease of identifying the reviewer. Using software to detect plagiarism will bring up published material that shows a similar style to the ‘anonymous’ reviews. We all have distinct styles of writing.
Determining who did the research may often be possible, but it would at least require the reviewers to read some of the paper. Would there be any harm in double blind review?
I am presenting this retracted paper in my lab meeting this Friday. I’ll let you know what we think.
Update: Utrecht University says Dhonukshe manipulated data in four articles. Here’s the article in university newspaper DUB, in Dutch http://bit.ly/1gz8QuW and English (Gtranslate): http://bit.ly/NtU0gI
Just a few weeks back when all this happened, PD interviewed me on Akype and didn’t select me for a postdoc position. This is probably the first time I’m feeling lucky on not being selected for a position.