How To Read A Retraction III

Feb 05 2009 Published by under Conduct of Science, Ethics

From today's Cell:

We realized that the anti-IKKa (IB) loading controls presented in Figures 3A, 3B, 3C, and 4C are duplicate presentations of the same gel lanes and do not represent the correct controls for the individual experiments. In addition, the anti-IKKa (IB) loading control in the right panel of Figure 4C is an inadvertent duplication of the DNA-PKcs (IB) data in the left panel of Figure 5F. These errors in figure preparation limit the interpretability of the related experimental data in these figures, which are an essential component of the support for the main conclusions of the paper regarding the activation of IKK and NF-kB. We are therefore retracting this paper. We apologize for these errors and for any inconvenience they may have caused. Despite these errors, we stand by the reproducibility of the experimental data and the conclusion, which has been reached by numerous subsequent studies, that IKK and NF-kB are required for activation of innate immunity.
Augusto Lois, a coauthor on the original manuscript, was not reachable via any of the available contact information and therefore has not seen or agreed to the text of this Retraction.

Realized!? Inadvertent!? Limit the interpretability!?
HAHAHAHAHAHAHAHAHAHAHAHAHAH!!!!!!!!!!!!!!!!!

73 responses so far

  • James F says:

    Don't you hate it when former coauthors skip town?

  • Becca says:

    "IKK and NF-kB are required for activation of innate immunity"
    Well, that part is true!

  • DrugMonkey says:

    Interesting. Only nine years later?
    Do you notice how they don't actually say the missing author was the one who faked the data? Yet that sure as heck is the impression generated, isn't it?
    Wonder if he was some poor schmuck undergrad.

  • neurolover says:

    "duplicate presentations of the same gel lanes"? How can that possibly take 9 years to discover? And, I ask naively as someone who never looks at gels. And, I presume this is a Cell paper, so we don't have the excuse that no one ever read the paper, like one might in an obscure journal.
    BTW --PP your link isn't working.

  • DrugMonkey says:

    well, neurolover, presumably anyone who tried to reproduce the data with the right controls was able to replicate the experimental finding, going by the robust assertion at the end of the retraction. Somehow I expect that the motivation to squawk is pretty low when the original finding is right, even if they've faked their control data.
    the real question is what motivated this retraction 9 yrs later? a larger scale investigation of fakery maybe? that might explain why one of the authors (who google suggests was at least a postdoc and has patents with the PI. if it's the same guy, more recent patents and LinkedIn associating him with a private company in the same town as the PIs lab. if he's really disappeared from contact....) is supposedly AWOL

  • S. Rivlin says:

    Based on some experience dealing with fraudsters and their sneaky ways, DM's speculation on a larger scale investigation of fakery sounds very plausible. It possibly came to light only recently, forcing the retracting authors to recheck other experiments and papers on which said "disappearing" author was working on at the time. We may or may not hear about additonal retractions.

  • All right!!! Who's kidnapped Fucklington!?!?!?!?!?
    Two on-topic non-delusional-asshole comments in one day?

  • Isis says:

    I think that should have read "Lane 1 was accidently Photoshopped onto Lane 4. The asshole who did it could not be reached for comment because the jerk skipped town."

  • Ed Yong says:

    But it was an "inadvertent" duplication. Clearly, what happened was that someone had Photoshop open, tripped, and fell on exactly the right combination of keyboard shortcuts needed to duplicate the lane.

  • Mopuse says:

    Well this first author is an alumni from a very famous lab whose articles have been facing many questions and doubts lately. It turned out it is a pattern for people from that lab to erase and copy-paste their gel photos. They also used a same photo in multiple articles and name it for completely different experiments.

  • neurolover says:

    So, I don't know what gels look like, but I can see this happening in my field with images -- because the pattern recognition of complicated images can be poor enough that if you're creating a composite figure you might copy an image from the wrong file and not notice. But, if that happened "accidentally" it would be grounds for a corrrigendum, right? Replace the figure with the right one.
    I guess you then argue that you can't do that because now, 9 years down the road, you don't have the right gel anymore (documented, saved in the right place and all that)
    I hate this stuff. I hate it when people try to argue accidental error when there was really fraud ('cause it makes the cases of accidental error, which do really occur) difficult. I hate it when people blame underlings. I think if you're the corresponding author you signed off on the paper, the same way we now require CEOs to sign off on corporate reports, and that you are liable. It then becomes your responsibility to make sure that you know enough to sign off (or if you can't, that you accept the risk).
    Oh, and the other authors are liable, too. I'm not sure I'd make them jointly and severally liable, but I'd make them liable enough that people would think twice about signing on to author lists without knowing what they're doing.

  • cashmoney says:

    a very famous lab whose articles have been facing many questions and doubts lately.
    Do you have published figures you can point us to or are you rumour-mongering?

  • Julie Stahlhut says:

    This piqued my curiosity on a slow day in the lab, so I've pulled up the gels from the original paper. The figures described as duplicates are not identical, although a few individual bands or small groups of bands appear identical between at least two of the figures. So, if there are "duplicate presentations of the same gel lanes", they are probably heavily edited. Yikes!

  • neurolover says:

    Yeah, I just got piqued enough to look at the original. The retraction description had me thinking they'd just duplicated one image (the kind of thing that could happened if you had each of your controls labeled in sequence, but accidentally opened the same one each time when making the composite figure, and then didn't notice). But, as Julie says, only parts of the gels in question seem replicated. The replication would not strike someone's eye without unless you were looking at it very carefully.
    (though, it's to be noted that only the duplication in Figure 4C is called an "inadvertent" duplication in the retraction).
    Again, I'm seriously freaked out by this, and it's not my field. I think the argument that the problem would have been discovered earlier if the conclusion was not actually true is valid (and, I'm guessing that becca's statement that it's true means that there's really been independent verification in the field.). But, how often do we meet that standard -- of independent verification of the main points of a study? There are many fields in which that kind of replication is very rare.
    I am particularly fearful of arguments that distribute the responsibility so that no one is ultimately responsible for the screw-up (or they assign the responsibility to the third mail clerk form the end, the way the bankers would like to).

  • Eskimo says:

    Even eight years ago, nobody needed to do a study asking whether "IKK and NF-kB are required for activation of innate immunity."
    The point of the original paper seemed to be that DNA-PKcs (the gene mutated in scid mice) was involved in sensing the ends of bacterial DNA for innate immunity.
    Didn't one of the Toll-like receptors turn out to be more important for sensing bacterial DNA?

  • JSinger says:

    Do you notice how they don't actually say the missing author was the one who faked the data? Yet that sure as heck is the impression generated, isn't it?
    I had the same reaction, but that's not really fair, is it? There's nothing suspicious about 1) not being able to find some old grad student (I've had the opposite problem, where some mystery guy deserved authorship and couldn't be found) and 2) noting that an author didn't participate in the retraction for whatever reason.

  • Becca says:

    "Didn't one of the Toll-like receptors turn out to be more important for sensing bacterial DNA?"
    TLR9 actual controversy, none of this silly "maybe IKK and NF-kB aren't required for innate immunity!" nonsense.
    There are other innate immune system components that recognize various nucleic acids typically associated with pathogens. I'd not venture to state which are more important as a blanket statement, but TLR9 better studied than a lot of them.

  • Anonymous says:

    It seem that MK's lab has the tradition. It has been exposed several month ago by another bbs forum. If you don't believe, just look at some other papers published on nature or science before 2006. Use Photoshop to zoom in every gel figure, I am sure you will find the trace.
    And look at another paper by the first author, Imunity 1999, fig 1e and fig6

  • neurolover says:

    So, what's the scoop about whether the conclusions of the paper are valid or not? I'm interpreting the chit chat to say that they carefully stated the conclusion as something that is well recognized, but that the paper itself was arguing something else. True?
    Yeah, this kind of thing makes me feel kind of like catholics must feel about pedophile priests, like my religion has been tainted (you know, aside from the seriousness of the transgressions themselves).

  • DrugMonkey says:

    I am far from expert but the problem seems to be with the controls. Faked controls could mean that the "real" experimental data are artifact...or not. My point was that the protestation in the retraction is that the findings have been subsequently confirmed. So their critical experimental could have been correct, it is just not demonstrated in the accepted way by those figures.
    In my parlance, suppose you gave a bunny an injection of amphetamine and it hopped higher that you ever observed it to hop untreated. So you conclude amphetamine enhances hopping. ...buuuut, you didn't do the vehicle injection control. so it could be the case that you are seeing an arousal response to being held and stuck with a needle, rather than a drug effect. so you haven't really demonstrated the effect. Now suppose a bajillion other labs replicate and extend the result. You were right from a factual standpoint but you did not demonstrate the reality with your missing-control experiment.

  • S. Rivlin says:

    DM, your "scenario" is somewhat wishfull thinking. Maybe in an ideal world of scientific research, such a scenario would be the standard and we would see many more papers being retracted for these kind of reasons. However, in the real world of scientific research performed, manipulated, adjusted, falcified and fabricated by unethical scientists, a more realistic reason for retraction nine years after publication has to be due to something much more sinister than "drug test that lacked a vehicle test as control."

  • DrugMonkey says:

    Sol, all I'm saying is that I do not have the background in the field, nor the inclination to read the paper closely enough, to offer an opinion on the scientific nitty-gritty. I am going by assertion of the retraction that the faked data were loading controls and that the experimental data are reproducible.
    that is the conversation neurolover and I are having.
    why they are getting busted at this relatively late date and bothering to issue a retraction is a very interesting, but different, question. Me, I am anticipating that this is part of a larger investigation because I am a cynic when it comes to this stuff.

  • S. Rivlin says:

    DM, I knew that we have an agreement somewhere. 😉

  • Becca says:

    "So, what's the scoop about whether the conclusions of the paper are valid or not?"
    Figure 3 is a real yawner- nothing not seen a million other times. If for some reason your experimental DNA isn't activating IKK and NF-kB (and you haven't like, mutated MyD88 or something wonky), there is probably something wrong with your DNA.
    If you manage to get significant levels of IL-6 or (especially!) IL-12 out of BMDM cells without activating NF-kB, tell me how. In Nature, most likely.
    So all of those conclusions would be solid. The wonky part of Figure 4 (c) just shows that the IKK activation by their DNA is blocked by Wortmannin, which (again) has been seen eleventy billion times.
    In short, they retracted the paper without needing to retract any conclusions. Very weird stuff.
    The main novelty of their paper, the DNA-PK stuff, is the stuff I'd be more inclined to doubt. Particularly since a Pubmed of "DNA-PK MyD88" turns up zero papers, and the bacterial DNA-> MyD88->NF-kB signaling pathway is incredibly well established.

  • Jimmy says:

    another interesting aspect of this story. MK (Michael Karin) was a corresponding author on the 2000 Cell paper but now is not a corresponding author on the retraction. Why's this?
    Another thing. Did you guys notice that in the previous issue of Cell, MK published an erratum for the Budanov and Karin paper for similar shady image processing irregularities. Why does th Budanov and Karin paper get to be corrected but the other paper retracted?
    Also: you guys should check out the paper by Zhang, Karin, Chu and colleagues in EMBO Reports, 2005 "Activation of IKK by thymosin 1 requires the TRAF6 signalling pathway". Tell me if I'm wrong, but doesn't the anti-IKK blot in Figure 2B look to be the copied directly from the Chu et al 2000 Cell paper that was just retracted? Compare anti-IKK blots from EMBO reports Figure 2B paper to the 2000 Cell paper Figure 3A. I've run 100s of gels in my lifetime and bands never come out looking as identical as they do in the gels published in EMBO reports and Cell.
    I suspect something much bigger is going on here but only time will tell.

  • Mike_F says:

    Jimmy - you should forward that query to EMBO Reports. Would be interesting to see how they deal with the problem... . Mike

  • Jimmy says:

    Not sure how to forward this concern to EMBO. Suggestions?

  • mark says:

    you are right, that is the exact same blot! Atleast run another gel with the same proteins to make it look like a little different. If you are going to cheat, dont be lazy about it.
    Just awful.

  • Jimmy says:

    would someone like to volunteer to send a note to the EMBO editor about the similarity in the anti-IKK blot in Figure 2B of the Embo paper and the Chu et al 2000 Cell paper that was just retracted? Compare anti-IKK blots from EMBO reports Figure 2B paper to the 2000 Cell paper Figure 3A.

  • whimple says:

    Just email the editors anonymously Jimmy. Use a disposable gmail account.

  • Jarvis the dancing monkey (usually posts under another name) says:

    I just sent an email to the EMBO senior editor pointing him to this page, suggesting he specifically read comment #26 and the ones that immediately follow. I used my real email & contact info.
    I think it's important to bring these things to light when we notice them. Good eye, Jimmy!

  • Jimmy says:

    have you heard anything from the EMBO senior editor about this?

  • Jarvis the dancing monkey (usually posts under another name) says:

    I've heard nothing at all, Jimmy. Others should feel free to write. Editors are busy people. They don't necessarily have time to surf the internet because some stranger said they should, especially when the reward is finding out their journal screwed up.

  • Dr. Feelgood says:

    As a reviewer I can recall at least two papers I have reviewed that I would consider scientific misconduct cases. Both within the last 2 years. One involved a chinese group that plagiarized several paragraphs word for word from a paper they cited. When I went to look at the cited paper, I was floored. It was hilarious. The editor noted this to the authors. I never saw it again. (Maybe forgivable due to language barrier? Dunno) The second paper was one where they tried to push old data as new. Five of the six data figures were the exact same data as a previous paper from the year before with an extra group added in. When this was brought to the attention of the editor, he advised the authors. They said that they hadnt "realized" this, and that they would revise and credit the prior paper. Never saw that one again either.
    My problem is that when you bring these issues to an editor, they are often so milktoasty and mealymouthed so as to not offend authors. My view is, forward the notification off to their department chairs or deans of research as a carbon copy. Let them give their explanations to those who pay their salaries.
    Its amazing how much pseudo-fraud like the kind I mentioned gets through. It makes me worry how much out and out fabrication there is out there. I hardly ever believe anything I read anymore. If its interesting and related to what I am doing, I wont cite it until I can reproduce it, or have some experimental basis to believe it.
    DM/CPP remember the good old days of your youth when you read a paper and just believed it because it was printed? Hehehehe....That ended in year two of grad school....*sigh*
    Doc F

  • Jarvis the dancing monkey (usually posts under another name) says:

    I heard from the editor(s). They are looking into it. The editor did note, however, "that Fig 2B, that is under discussion in the blog in terms of its similarity to Fig 3C of the Cell paper, is not the correct version, since the authors published a corrigendum to this paper in 2006: see EMBOR 7 (10) 1059."
    So maybe some benefit of the doubt is due here. The editor promised to be back in touch once they have looked into it more deeply.

  • Jimmy says:

    ok, so this is really suspicious. Figure 2B was changed in the corrigendum with the comment:
    "It has recently come to our attention that Fig 2B is incorrect. The correct figure appears here. This change has no effect on our conclusion that IKKbeta is required for activation of IKK and NF-kappaB by thymosin alpha1."
    What does that mean that Fig 2B is "incorrect"? Why would they use the same gel from a 2000 Cell paper (that is now retracted) in a 2005 EMBO reports paper?
    It looks like they just put in a new blot once they were caught using their old data.
    I hope these guys aren't able to get away with this. There are lots of other examples of suspicious data manipulation.
    For example: take a look at another paper from Michael Karin's lab:
    Luo, J.L., Tan, W., Ricono, J.M., Korchynskyi, O., Zhang, M., Gonias, S.L., Cheresh, D.A., and Karin, M. (2007). Nuclear cytokine-activated IKKalpha controls prostate cancer metastasis by repressing Maspin. Nature 446, 690-694.
    Check out Figure 2a. Kai1/Cd82 blot. Lanes 5-7 appear to be identical to lanes 2-4. Isn't that strange?
    Also, take a look at this other paper:
    Briata, P., Forcales, S.V., Ponassi, M., Corte, G., Chen, C.Y., Karin, M., Puri, P.L., and Gherzi, R. (2005). p38-dependent phosphorylation of the mRNA decay-promoting factor KSRP controls the stability of select myogenic transcripts. Mol Cell 20, 891-903.
    Check out Figure 1E, GAPDH blot. Lanes 5-8 appear identical to lanes 9-12. Notice how the shape of the band in lane 5 looks the same as that in lane 9.
    In the same paper, check out Figure 2C, GAPDH blot. You have to use photoshop to adjust levels to notice this, but it appears that the two gels were spliced together to appear to be the same gel.
    And same paper...
    Figure 3B, GAPDH blot. Lanes 9-14 appear to be identical to lanes 3-8. And even more remarkable, lanes 1-8 appear to be identical to lanes 1-8 in Fig. 1E.
    What is going on here?

  • Jimmy says:

    Sorry, I clicked "post" too soon. I found others:
    Check out this paper from Michael Karin's lab:
    Matsuzawa, A., Tseng, P.H., Vallabhapurapu, S., Luo, J.L., Zhang, W., Wang, H., Vignali, D.A., Gallagher, E., and Karin, M. (2008). Essential cytoplasmic translocation of a cytokine receptor-assembled signaling complex. Science 321, 663-668.
    Look at figure 1A, JNK blots (8th row down). For the TRAF3-/- and TRAF6-/- samples, the gel bands appear to have been duplicated. Next, take a look at figure 6D. The 7th row down (IB:Tak1), it also looks like gel bands have been duplicated (- and SM blots look identical).
    Thoughts?

  • whimple says:

    corrigendum added to: Nature 446, 690-694
    Mol Cell 20, 891-903. Check out Figure 1E, GAPDH blot. Lanes 5-8 appear identical to lanes 9-12
    Not identical at the individual pixel level.
    In the same paper, check out Figure 2C, GAPDH blot. You have to use photoshop to adjust levels to notice this, but it appears that the two gels were spliced together to appear to be the same gel.
    I don't see this.
    Figure 3B, GAPDH blot. Lanes 9-14 appear to be identical to lanes 3-8.
    Not identical at the individual pixel level.
    Science 321, 663-668.
    Look at figure 1A, JNK blots (8th row down). For the TRAF3-/- and TRAF6-/- samples, the gel bands appear to have been duplicated

    Not identical at the individual pixel level.
    Thoughts? Unless someone gives up a confession, this isn't going to go anywhere.

  • Spider says:

    I think there are two situations which may be worth you considering. When complaints are made about a paper post peer review and publication they are frequently evaluated by the corresponding author. If they go through the data images postpublication and discover a data error postpress then they may ask the journal to publish a correction. If the corresponding author believes the data to be compromised then they will usually contact the journal for a retraction. In the above cases the PIs who contacted the journals for corrections or retractions are probably blameless, they aren't trying to cover up mistakes or fraud, in fact are trying to sort out a papers issues. If the journal itself has to force a retraction as has happened then that is another matter in my opinion and in this situation I think it casts a shadow over the PI concerned.

  • jarvis the dancing monkey (usually posts under another name) says:

    Matsuzawa, A., Tseng, P.H., Vallabhapurapu, S., Luo, J.L., Zhang, W., Wang, H., Vignali, D.A., Gallagher, E., and Karin, M. (2008). Essential cytoplasmic translocation of a cytokine receptor-assembled signaling complex. Science 321, 663-668.
    Look at figure 1A, JNK blots (8th row down). For the TRAF3-/- and TRAF6-/- samples, the gel bands appear to have been duplicated. Next, take a look at figure 6D. The 7th row down (IB:Tak1), it also looks like gel bands have been duplicated (- and SM blots look identical).

    I looked at this one because it's recent and in Science. It's the only one of Jimmy's examples that I checked out. I agree the images look really very similar, but I put it into Photoshop, blew it up, cut, overlaid, and compared. While the images are very similar, there are clear differences at the single pixel level. This does not mean they aren't pictures of the same bands, of course. It just means it's unlikely to be the same picture.
    I think it's great that Jimmy is voicing suspicions; I applaud his skepticism and discerning eye. But it's important to tread carefully. Re-use of gel combs and idiosyncrasies of cameras can also make different gel images look like they might be the same thing when they are actually independent experiments.

  • I think it's great that Jimmy is voicing suspicions; I applaud his skepticism and discerning eye. But it's important to tread carefully. Re-use of gel combs and idiosyncrasies of cameras can also make different gel images look like they might be the same thing when they are actually independent experiments.

    In this regard, when I was a post-doc I frequently used an x-ray film processor that would always leave an identical pattern of little black dots on every piece of film. I never figured out why this was, but seeing these kinds of artifacts always in the same place on images of blots could look suspicious.

  • theorist says:

    I think the Matsuzawa paper is simply a case of a mistake that was corrected.
    What some of the guys here also don't seem to realize is that the journal itself frequently generates errors when taking a submitted paper and turning it into print form, and they also alter the images from the authors into the print format of the journal using their publish software applications. This is how a lot of mistakes and errors crop up in publications.

  • Jimmy says:

    perhaps you are right. i don't know. of course mistakes happen but it just seems odd to me that there are several of these high profile papers from the same lab with these types of 'errors'. And one of the papers, the Cell paper, was just retracted. Maybe there's nothing to it more than honest mistakes, but it just seems fishy to me.

  • theorist says:

    That PI has, if you check medline, 414 publications. Many are through collaborations and they have also a very large lab of international postdocs who stay for only a few years and move on. I think it just impossible to keep track of everything and mistakes get made. Sure the Cell retraction may look bad to some, but check the 1st and corresponding authors affiliation at that time it isn't the same as the PIs and most likely the final blame for the data duplications lie there i think. Still they did the right thing and pull the paper based on the errors in the control data.

  • jimmy says:

    " I think it just impossible to keep track of everything and mistakes get made."
    you're kidding, right?
    fool me once, shame on you. fool me twice... can't get fooled again.

  • qaz says:

    Re #46 and #47 - " I think it just impossible to keep track of everything and mistakes get made."
    I assume this PI has lots of NIH $. If so, then should this PI be getting NIH money if he is trying to run a lab that is too big that he can't keep track of everything and is making mistakes? Wouldn't it be better to fund 10 smaller labs actually able to do real science instead of sloppy crap?

  • Young Love says:

    I assume this PI has lots of NIH $. If so, then should this PI be getting NIH money if he is trying to run a lab that is too big that he can't keep track of everything and is making mistakes? Wouldn't it be better to fund 10 smaller labs actually able to do real science instead of sloppy crap?

    This PI has seven R01s and a couple other grants. It's obviously easy to write CNS papers and proposals when you're not hobbled by detail.

  • theorist says:

    Well this is the point you don't seem to understand qaz, that the retracted paper was a collaboration between four different affiliations. The PI we're chatting about is famous for making transgenics, but also frequently gives materials from these to collaborators. I think the ultimate culpability lies with the first author, who was also corresponding author (also a PI, but less famous) as they probably generated most of the actual figure data.
    You also have to bear in mind generally that going over every paper with the latest photoshop at the pixel level looking for duplications is a relatively new phenomena in research. Loading controls frequently look very similar so inadvertent mistakes are pretty easy to make and hard for researchers to spot without going through everything in hi zoom. I don't think many PIs do check their postdocs or collaborators figures at the pixel level for mistakes. Perhaps we just live in more cynical times where people focus on 1 bad collaborative paper and ignore the hundreds of important contributions to research...

  • Jimmy says:

    I think the point is that this does not seem to be 1 bad collaborative paper, but rather a string of high profile CNS papers that have questionable image processing irregularities, one of which has been retracted.

  • theorist says:

    I'd leave it up to the journals in question Jimmy. As it stands we are just talking one retraction of a paper 9 years old. Everything else seems to just be a few papers (2-3) simple mistakes which the actual journals and editors have agreed to. Remember the journals can retract a paper without author consent if they believe it breaches their policies on publication.

  • qaz says:

    Theorist #50 -
    I make no claims about the validity of the work that this PI is doing. It is very far out of my field and I am not qualified to judge it. What I was saying (which I stand by 100%) was that anon #46 seemed to be defending this person because his lab was too big to watch carefully. I do not believe NIH should be funding a lab that makes inadvertent mistakes because the lab is too big to watch over. To echo a fight that Dave and I had a while ago about ICR - there is only so much NIH $ to go around and seeing waste like this makes me mad. A PI who makes stupid mistakes is a waste of NIH money that could be better spent on the small struggling labs with PIs who have the time to get it right.
    Personally, I do not think there are a lot of true frauds (or outright criminals, as S. Rivlin correctly defines them). On the other hand, I think there are a lot of big empires sucking money from real science. Some empires can be good. But this empire apparently includes a PI who lets stupid mistakes slip through and does not have time to do the science right.

  • Cell Signaling Wizard says:

    theorist,
    Karin has been "correcting" and/or retracting a lot of papers recently. He just had a correction on duplication of bands in Science as well as in Cell (Sestrin paper, not this retracted one). I'm sure that this isn't just a one time thing. In fact, I'm sure of it because... Well, we'll leave it at that.
    BTW, Karin has 20+ post docs and is known to be very aggressive. The combination of those two factors equate to very bad things.

  • theorist says:

    I think it's only one retraction and some corrections, the other one you mention does look pretty sloppy work. It is a complicated situation if you have quite senior scientists in a large lab more or less writing papers themselves and making mistakes, but I can see no whole scale attempt at fraud here at all.

  • Experimentalist says:

    It's so frustrating to see that people can easily get away with this kind of crime. I wish there are a better punishments for those who did this. It's also frustrating to think that these are the kind of people who become future faculty. It totally put me off of science.
    I guess I should commend the PI for retracting the paper even after 9 years. Some PIs in this situation will not agree to such retraction because of the many consequences that all of us are aware of.

  • billope says:

    This series of posts seems dead, but I just found it after I noticed the Cell retraction. I heard that Dr. Chu was denied tenure at Brown, but I'm not sure if this retraction had anything to do with it. Has anyone heard? I suppose he was not required to give back his R01 that was based on the Cell data.
    As of today EMBO Reports has made no change to the paper noted above that had the figure from the Cell paper.
    I looked through a few of Dr. Chu's other papers and found that the following articles contain obvious duplicate figures. Moreover the figures are in some cases labelled differently and are said to be in different cell lines.
    JBC:
    J Biol Chem. 2008 Oct 24;283(43):28897-908. Epub 2008 Aug 20.
    AMP-activated Protein Kinase Contributes to UV- and H2O2-induced Apoptosis in Human Skin Keratinocytes :
    Cong Cao, Shan Lu, Rebecca Kivlin, Brittany Wallin,Elizabeth Card, Andrew Bagdasarian, Tyrone Tamakloe, Wen-ming Chu, Kun-liang Guan and
    Yinsheng Wan.
    SEE ESPECIALLY Fig 2G and 2H (Matches Fig 5A, B, and C in the Cell Signal paper below)
    Cell Signal:
    Cell Signal. 2008 Oct;20(10):1830-8. Epub 2008 Jun 24.
    EGFR activation confers protections against UV-induced apoptosis in cultured mouse skin dendritic cells.
    Cao C, Lu S, Jiang Q, Wang WJ, Song X, Kivlin R, Wallin B, Bagdasarian A, Tamakloe T, Chu WM, Marshall J, Kouttab N, Xu A, Wan Y.
    SEE ESPECIALLY Fig 5A, B, and C (Matches Fig 2G and 2H in JBC above)
    There may be other duplications I did not find. It was not hard to see these.
    Very disturbing indeed! Now I see with Jaundiced eye that Dr. Chu has a Science Signalling paper....

  • patternist says:

    In my informal examination of publications by Drs. Wen-Ming Chu and Yinsheng Wan, I also noticed the following articles with obvious duplicate figures that no one has mentioned yet. Again the figures are in some cases labelled differently and are said to be different treatments or cell lines.
    Science Signaling Fig 6D (Gi/Go DKO;36hours) matches Fig4D (CaOV3; -/- treated cells) in Can Lett paper.
    Science Signaling Fig 6D (36H WT) matches Fig 4D (CaOV3: +/+ treated cells) in Can Lett paper.
    Science Signaling Fig 6D (FBS; WT) matches Fig 4D (CaOV3; +/- treated cells) in Can Lett Paper.
    [C. Cao, X. Huang, Y. Han, Y. Wan, L. Birnbaumer, G.-S. Feng, J. Marshall, M. Jiang, W.-M. Chu. Gai1 and Gai3 are required for epidermal growth factor–mediated activation of the Akt-mTORC1 pathway. Sci. Signal. 2, ra17 (2009).]
    [Cong Cao, Shan Lu, Alex Sowa, Rebecca Kivlin, Ashley Amaral, Wenming Chu, Hui Yang, Wen Di, Yinsheng Wan. Priming with EGFR tyrosine kinase inhibitor and EGF sensitizes ovarian cancer cells to respond to chemotherapeutical drugs. Cancer Letters, 266, Issue 2, Pages 249-262]

  • DK says:

    Bwa-ha-ha! I totally missed this retraction. Obviously, the retraction is just as dishonest as the original BS was. Gosh, there is after all *some* justice in the world. Goo.
    A personal anecdote: Many, many years ago I interviewed with MK for a postdoc. Got an offer, refused. MK's lab, tremendously successful it was by every mean, left a sleazy impression. I could never explain it, it was just a feeling. Not being around that particular field, I had no idea how it is doing and all that. But boy, I am soooo not surprised by the retraction!

  • Whatson says:

    If this stuff regarding the other Chu papers is true, is anything happening with them? That looks really bad! Does he still have a lab? Funding?
    Is this how science really works?

  • pipettor says:

    Check out this!! Looks like this guy Wen-ming Chu found a job at the University of Hawaii Cancer Research Center. How could they have not checked into his background? Unbelievable....
    http://www.starbulletin.com/news/20100331_3_specialists_join_UHs_Cancer_Research_Center.html

  • pipettor,
    Does it occur to you that this may have been investigated and Wen-ming Chu actions found to be inadvertent. Just because a paper is retracted or changed doesn't mean it is always deliberate fraud. Perhaps it just a sign o' the times that folks are so cynical.

  • HesProbablyNotInnocent says:

    "Does it occur to you that this may have been investigated and Wen-ming Chu actions found to be inadvertent. Just because a paper is retracted or changed doesn't mean it is always deliberate fraud. Perhaps it just a sign o' the times that folks are so cynical."
    Maybe if it was just one paper retracted but there are multiple recent papers by the Karin lab that are in question and other papers by Wen-ming Chu that are also suspect. Read the above thread and you will see the data for yourself.

  • patternist says:

    Well, now the Science Signaling paper listed above has been "corrected". I hope he pulled the right files THIS time. Not sure how we know.
    The JBC paper was WITHDRAWN. Which I believe is the appropriate thing to do.
    No word on the EMBO Reports, Cancer Letters, or Cell Signaling papers....yet?
    And any grants based on this work?
    And does anyone know whether an investigation led to Dr. Chu leaving Brown? Perhaps these changes were not found to be "inadvertent" and tenure review offered a way to quietly resolve what could have been an otherwise difficult situation.
    All told, this seems to be an ongoing and sad series of events. One of several similar ones lately...

  • question says:

    So, is NIH going to continue funding Dr Chu's research ?. The retraction is for work published in 2000, likely the time when he started to apply for R01s?.
    NIH peer review is based on competition. Dr Chu's brilliance might have put out of business some other scientists working equally hard but with much higher sense of responsibility....

  • Physician Scientist says:

    OMG

    This grant was RENEWED!!!!

    Project Number: 5R01AI054128-07 Contact PI / Project Leader: CHU, WEN-MING
    Title: MECHANSIM OF ACTIVATION OF INNATE IMMUNITY BY ISS-DNA Awardee Organization: UNIVERSITY OF HAWAII AT MANOA

  • DrugMonkey says:

    If I'm reading the comments correctly, this guy has retracted and corrected papers but never been found guilty of fraud? I know what you mean but we do have commitment to "innocent til proven guilty" in the US....

  • dave bridges says:

    innocent until proven guilty, but im sure reviewers sure can take into account historical accuracy of a lab. Better to renew a grant of a good lab with an instance sloppy record keeping (if thats the case) than a non-retraction lab whose data is never reproducible

  • [...] Physician Scientist notes on a prior post that an individual scientist under suspicion for several dubious papers has retained [...]

  • Spiny Norman says:

    Whimple -- the blots are not expected to be identical at individual pixel level because they are different PDFs which have been subjected to separate JPEG compressions, almost certainly using different compression parameters and implementations of a JPEG compression engine.

  • Correct Science says:

    http://www.plosbiology.org/article/info:doi/10.1371/journal.pbio.1001848
    Another correction for Cao Cong's record: this time a hidden COI for John Marshall. Looks like it was imposed by the editors, given the copyright statement.

    http://news.brown.edu/pressreleases/2013/02/angelman
    Wow, the senior author was inventor of the drug on two patents, owned a Biotech company called Angelus Therapeutics and was actively trying to commercialize the drug he was reporting whilst claiming no conflict of interest. Why is this not retracted? Seems too little too late.

  • Michael Greenberg says:

    This is an old thread but an expression of concern has now been published for Cao Cong, Dennis Goebel and John Marshall on this PLOS Biology paper. There are concerns related to more duplicated figures, including duplicated bands at different exposures: 3 duplicates in one figure alone. In addition, vertical resizing of duplicates that change their appearance. Many other issues.

    https://pubpeer.com/publications/FD2CF905EF612DC4B9162D9F482D73

    Now, concerns over another Cao Cong, Marshall paper figures.

    https://pubpeer.com/publications/1A64630EE2935B4F076A7EE693AD56

    You would think that after all of the above retractions and corrections, they would have learned their lesson. Although... perhaps they did... just not the lesson we would have hoped. Cao Cong is now a Professor in China!

  • Mike says:

    https://pubpeer.com/publications/BC2BDE850086C80B0946E0948184AF#fb49929

    More shoes are starting to drop on Cao Cong and John Marshall papers.

Leave a Reply