The Journal of Biological Chemistry has posted another of its inscrutable and opaque retraction notices, this one for a study first published in September 2011. The retraction reads in full:
This article has been withdrawn by the authors.
In the paper entitled “Isd11p protein activates the mitochondrial cysteine desulfurase Nfs1p protein,” Debkumar Pain, associate professor at the University of Medicine and Dentistry of New Jersey, and associates found two proteins that were inactive on their own, but when bound together acted like scissors to snip of sulfur atoms from the amino acid cysteine.
Pain and company retracted the paper, and when contacted by Retraction Watch, he added to the mystery:
The original data as published in the JBC paper are correct, and the conclusions are valid and fully supported by those original data. There was no falsification of any data. To respond to your queries point by point, I must obtain permission from the Research Administration here.
So, according to the senior author, the paper has valid results and conclusions, yet the publication got yanked.
Hmm.
We also asked the journal the story behind the retraction, and we got this message emailed from Nancy Rodnan, director of publications at the American Society for Biochemistry and Molecular Biology, which publishes the JBC:
I understand you have contacted ASBMB staff regarding a retraction posted in JBC requesting information. As I have stated several times before, to Retraction Watch, this information is confidential. My suggestion would be to contact the corresponding author which you have done, and wait for a report from the author when he has permission from his institution.
We obviously disagree about whether the reason for a retraction is confidential. So does the Committee on Publication Ethics, which recommends that retraction notices should
state the reason(s) for retraction (to distinguish misconduct from honest error)
Two weeks after our initial request, Pain, who seems to be willing to talk to us, hasn’t gotten the green light from his research institute.
I must follow the University policy and I will get back to you when I get the green light. At this time, I have no idea how long it might take. Thanks for your understanding.
Given that UMDNJ is a public institution, we’ll file a FOIA request if need be, but we hope that won’t be necessary. We also hope the university isn’t stonewalling just to keep us from writing about this retraction.
We’ll keep you posted on what we find out.
Interesting story. In such cases a reasonably likely possibility is that by retracting the paper the authors can avoid being found guilty of misconduct. There has clearly been some form of formal institutional procedure at UMDNJ, so it seems that no one considered that people might actually want to know the facts.
“original data” x 2
Me thinks that there is some recycled data in the paper.
Since when does a PI need the permission of his school to respond to a question about one of his papers?
JTFC, this stuff isn’t even funny any more…
Figure 1. A. MW marker bands at 66 and 14 kDa are duplicated. Also band in lane 1 (Nfs1p) and lane 3 (NifS) are duplicated – just blow up the resolution to see it!
Figure 2A. Uniform gray background. The band in lane 3 (250) has a shadow extending vertically above it, which abrubtly ends with a horizontal cut off. The exact same trick is evident in Figure 3A, only this time below the band (heck, it even looks like they may have used the same band!) Figure 3C, lanes 2 band also has a very odd shadow above and to the right of it, possible evidence something has been erased here?
Figure 4. Panels B and C show some nice fuzzy normal looking bands, with feint shadow above and below. There are some bands in these blots which have clearly been drawn in – the lines are very thin and have razor sharp edges, and no shadows above or below.
Figure 6A. Top left band has an absolutely flat top. Not a slightly wiggly top, consistent with normal gel bands and the bottom of the same band, but an absolutely 100% straight line cut off from the black of te band to the gray of the background. In drawing this cut-off, it looks like they missed a bit over on the right side of the band.
Figure 6C – the first last is pasted in. I could see this from across the room. lane #9 over on the far right is also pasted in, you can see the seams.
There are probably more “similarities” between bands in the different figures, but I don’t have my multiple monitors set up here, so can’t open all the Figures at once to see this.
The issue here is what’s best for ASBMB? Trying to save face and be as impartial as possible, and coming out looking like idiots to everyone who has a computer and knows how ot use CRTL + scroll wheel, or and growing a pair and being open about misconduct? Your choice ASBMB, but I think more than a few membership renewals (mine included) will be riding on how you handle this stuff in future. Who wants to be a member of a society that stares misconduct in the face and looks the other way?
There are some great Photoshop tools for finding image manipulation. They are available on the Office of Research Integrity website. Just plug them in and let the automation do a lot of the work for you.
Yikes. You have very good eyes, vhedwig… but I see what you see. “There was no falsification of any data.” Let’s see an explanation for this.
Thanks again to vhedwig for pointing out the anomalies in the figures.
Others may feel a sensation of standing on the edge of a cliff, with a stiff wind blowing in your face, and looking out, beyond the edge, to a yawning canyon. The data just blows away and you’re left with nothing but air to support your conclusions.
Why produce fake gel pictures if you have good data? How can they say their data is good when their pictures are phony?
How long do they think they can keep their motives for this retraction secret?
–puzzled monkey
Great question – one I have pondered many times with analogous stories. In the end I think all one can or needs to say is that scientific misconduct is just another form of criminality and hence the answer to this question is simply the whole field of criminal psychology.
Once I accepted this, then everything started to make sense. These people are not behaving like scientists but like criminals. Writing to them is a waste of time – they don’t care about integrity, good data or any of these issues because they don’t need to.
This attempt to hide the facts AFTER the retraction seems like insanity to scientists but is entirely consistent with criminology – how many trials do we hear about there the guilty party refuses to show remorse despite inescapable evidence of guilt.
Such people are absolute nightmares for institutions to deal with but I would have thought that the ORI would have made sure this case was sorted out. Instead we are left with a farcical outcome: retraction for no reason, obvious image fraud, yet authors who maintain that all is well!
it looks to me as they actually draw most bands by hand… that is a first!
Yeah, that was my first thought when I saw Figure 6B.
Is it worth noting that the PDF is stamped “withdrawn” instead of “retracted”? Is there a difference???
i think figures were made by some unskilled person, who made some silly mistakes. Otherwise, data looks real. I have high regards for JBC publications and if this paper was published there, it means paper went through rigorous reviews. Problems in figures are most common and many published papers has this kind of problem.
Are you new around here?
This is not okay, it happens in the best journals, and peer-review is not fool proof. Problems in figures are common on this site, but otherwise NOT common. This. Is. Not. Normal. Science. It should not be tolerated.
catherine: I fear that dismissing the alleged fabrications in this paper, as presented by vhedwig, with phrases like “unskilled person,” “silly mistakes,” “data *looks* real” (emphasis mine), and “[such] problems in figures are most common” suggests naivety that many will consider laughable, if not fantastic. The belief that publication by JBC “means [this] paper went through rigorous reviews” also merits reconsideration. Amongst the things to be learned from Retraction Watch is that the allegedly premier scientific journals are particularly prone to retractions and thus, by extension, oftentimes to reviews that are less than rigorous.
vhedwig is indeed right. This is image fraud.
One of the things RW has helped to reveal is how labs that commit fraud on one occasion will usually have done so on others (this is basic criminology: see http://www.retractionwatch.com/2012/04/05/jco-expresses-concern-over-western-blots-from-spanish-group-that-had-aroused-earlier-concern/).
So one might expect to find other papers from this group with problems. A quick search finds:
http://www.jbc.org/content/283/3/1362.full.pdf
Look at Fig.1A. That is a classic splicing artefact caused that should be assumed to represent photomanipulation.
Examining other papers by labs engaging in fraud is a critical responsibility of the research institution (although in this case the second paper is also in JBC so the journal could have picked this up). So one should expect at least an expression of concern over this second paper and potential further retractions from this lab. This may explain the radio silence from the authors.
Catherine – I agree with you at one level – the figures were made by an unskilled person. That is to say a person who is unskilled at image manipulation. A skilled fraudster would not have left such obvious clues that the gel images were fabricated.
Not sure on Fig. 1A, but would definately question 1B, 2B and 6A for a start.
Yep, I see 1A, B and 6A, also B. Someone tell the journal.
Just for clarity I’m referring to splice artefacts which is where the background of the gel suddenly changes in an artefactual way. There are also very thin white lines on most of these figures which I don’t quite understand but are not splice artefacts.
To me the problems are absolutely clear Fig. 1A (look at the top of the middle lane – clearly pasted on), 2B and the top panel of Fig. 6A . I’m not sure about Fig. 1B.
This lab has clearly been fabricating its gels for some time.
What is concerning is that neither the journal or the institution has commented on the wider problem, and one must assume they won’t unless encouraged to do so.
I’m in shock this morning. The quality of the manipulation in this paper is laughable – how the hell has this escaped notice for 4 years? Not only are the splice artefacts plainly visible, in most cases the borders of the cobbled together lanes don’t even line up – 1C is a beautiful example. My 6 year old nephew has better photoshop skills than this.
I have long given up on JBC as a high-quality journal. The slow but steady decrease of the IF of JBC suggests that I am not alone in holding this opinion.
Two good things about JBC are:
1: they provide the original pdfs submitted by the authors, and these often have figures with higher resolution than the fully formatted final version (so it is easier to spot the dodgy blots.)
2: They have quite strong guidelines to authors (similar to those from the Journal of Cell Biology).
(Unfortunately, unlike the Journal of Cell Biology, the JBC does not screen the figures in accepted manuscripts to check whether the authors comply…).
One has to remember JBC’s handling of the Gopal Kundu case. They refused to bow to threats and pressure from senior Indian scientists and steadfastly retracted the paper.
I would place JBC in a much higher category in terms of integrity than most other journals (certainly above Nature who talk the talk but don’t actually follow their own highly publicised policies).
There are also duplicated bands in this figure here: http://www.jbc.org/content/280/19/18604/F3.large.jpg
And if you look at Fig. 6C in the originally submitted version of the paper (http://www.jbc.org/content/early/2011/09/09/jbc.M111.288522.full.pdf+html) you will see that in addition to splicing, the band in lane 4 is identical to the band in lane 7. This does not seem to be naive cutting and pasting, it appears to be deliberate misrepresentation of the data.
Agree w/ amw. “Scientific misconduct is just another form of criminality.” We can’t successfully control the problems of misconduct unless we treat confirmed misconduct as a type of criminality. Sending a letter to the editor usually doesn’t do the trick, as the editor has a motive to cover up in hopes that no one will notice. This is why we need ORI with teeth and fully staffed all over the world.
Experience in business has shown that adequate rules, adequately enforced, are necessary to control business fraud; the same is true in science. Despite high motives, some people will be corrupted by the available rewards, and they won’t be nice about it.
So my friend has a new site, and I sent some images from Dr. Pain’s papers to her. It makes fro a pretty damning case…
http://www.science-fraud.org/?p=52
Shocking. In figure http://www.science-fraud.org/wp-content/uploads/2012/07/Pain12.jpg the last lanes in each image (65C) are actually duplications of the 5C bands. Easier to spot when looking at high resolution at http://www.jbc.org/content/280/19/18604/F3.large.jpg