Brutal honesty: Author takes to PubPeer to announce retraction — and tells us she’ll lose PhD, professorship

Eriko Suzuki
Eriko Suzuki

Over the past week, there have been a number of comments on PubPeer — a site of which we’re big fans — about a 2007 paper in Oncogene.

The comments suggested that the figures in the paper had problems. Some bands seemed to be duplicated, and one of the images looked very much like that of another paper.

Then, today, first author Eriko Suzuki left this comment:

As mentioned above, I agree that our paper published in Oncogene (Rituximab inhibits the constitutively activated PI3K-Akt pathway in B-NHL cell lines: involvement inchemosensitization to drug-induced apoptosis. Oncogene 26,6184-6193, 2007) contained some inappropriate figures. I just realized that I had misconducted pasting some JPG images as I had huge amounts of similar data at that time. According to these mistakes, I judge that the conclusion we drew in that paper cannot be fully supported. Therefore, I regretfully asked Oncogene Editorial office to retract our paper mentioned above. I must apologize that my misconduct has brought trouble to the journal, Oncogene.

But that wasn’t the end of Suzuki’s honesty. We contacted her to confirm that she in fact left the comment on PubPeer. She said she did. Then she said more:

I have already contacted the Oncogene editorial office and asked to retract the paper.

Since relevant data were included in my Ph.D thesis, my degree and position will be rescinded.

The position to which Suzuki refers is as assistant professor in the Department of Applied Biological Science at the Tokyo University of Agriculture and Technology. She earned her PhD at Keio University, also in Japan, in 2008, with the work in Oncogene completed at UCLA between 2005 and 2007. Her Oncogene co-authors are at UCLA.

Suzuki told us in another email that her university was not yet aware of the issue, but that she would “express my desire to apologize in a sincere manner to the authorities in a couple of days.”

The Oncogene paper has been cited 37 times, according to Thomson Scientific’s Web of Knowledge.

51 thoughts on “Brutal honesty: Author takes to PubPeer to announce retraction — and tells us she’ll lose PhD, professorship”

  1. I feel bad for her. It’s probably fair that she loses her position, but her degree should not be revoked unless it can be demonstrated that the misconduct was intentional.

    1. Yes, it would tend to thin the ranks of academia if everyone with errors in their PhD had it revoked. The uni I did mine at cannot revoke them because there is no mechanism to retract degrees. If she has an otherwise good publication record it would be unlikely that her institution has grounds to dismiss her, unless things work differently in Japan.

      1. I will never forget my advisor’s words on this issue: “it’s not important that you do not make a mistake in your thesis. It is important that you don’t say anything that is blatantly wrong.” I still laugh every time I think of it. With that comment, he told stories of the most famous historical figures in my field who had mistakes later found in their dissertation work.

        1. I think students make too much of a bear of the thesis. It’s a pass fail document. You don’t get a relative grade on it. And it’s very rarely searched by other researchers (meaning it’s not serving the field). It can give a home to half finished work that was not published, but what really matters are your journal articles. For a scientists, the whole fetish of humanities people spending months writing a thesis is silly.

          I honestly assembled mine in ten day spurt. Obviously most of the work was previously published in “real journals”. Yeah, thesis gave a place for a little bit of half finished work that was never published or not revolutionary enough (e.g. a structure confirmation). And a few comments on innovations in equipment or methods that are of benefit to the lab group, but maybe not worthy of a publication (built a doohickey in the machine shop, but not a whole new instrument).

          Yeah, there’s the lit survey and you could probably write a book (or several) if you really want to survey your field. But that’s where the comment about pass fail comes in. Do the minimum. If you do more, it ought to be a publication.

          I’m not in any way advocating cheating. If anything, I’m saying just be candid and honest (including when you dropped a sample on the floor or when you learned a lab safety insight). The main useful customer of your thesis is other students who follow you in your lab group. So, keep it real. As long as you have the regular journal publications and a job waiting, the professors won’t screw with you. You’ve done your part…get out of there.

          I’m not endorsing fraud (the opposite…be simple and direct versus hype-y), but I think what really speaks for you is your publications in terms of interaction with the broader community. How many people really check theses? They are infrequently cited, which implies people don’t rely on them.

          So if you have enough work in the publications, just cut and paste them into chapters and you’re almost done. Even the intro chapter…there’s different levels of effort possible in doing a survey of the field. You just need enough to satisfy the committee. The minimum! If you want to do some Greenwood and Earnshaw effort, fine…but that ought to be a journal publication. If you never bothered at the time, why do it now?

    1. Suzuki appears to be the consummate academic. I would trust her just for her searing honesty in this instance alone.

      As we all struggle through times of ‘excessive unbelievability’ among our politicians, lawyers and also our academic elite, is it not an outstanding act of integrity to admit that one has made a professional error? To state that one values one’s own honesty higher than one’s career prospects?

      Burnout? She makes a mistake and Sylvain Bernes judges her as BURNED OUT…??? Sylvain, please qualify the research that supports your statement. She has.

  2. The senior author of the paper has several potentially problematic papers on pubpeer where Suzuki is not an author.

    1. Are this honesty? Recgonize misconduct when you are caught and exposed? Never manipulate data must be consider honesty…

      1. There is a difference between making errors and dishonesty. What may be lacking in that lab is the importance of good record keeping. Plus a lot of pressure to publish, so there are a lot of different experiments being run, and a constant changing of research objectives. In the end it will probably be the significant results will probably be the ones produced by errors.

      1. This is why I think she went straight to public, appealing to the soft hearts of questioners. “Nothing more to see here, just let me be and go home”.

      2. The VC certainly has been well supported, and continues to be well supported by the National Cancer Institute.

        Take a good look at, and note the apparent irregularities, you may see in the link above in the VCs CV the first author and the VC have substantial grants, therefore presumably some of the data in the grant applications may have been from a similar, dubious source.

        Has the NCi any powers to look into this?

      3. While the Bonavida Oeuvre is the primary focus of that list, surely David you have noticed the historically interesting company being kept? For example

        Note also this equally exciting new entry at PubPeer since you posted that list

        Given your encyclopaedic knowledge of the dark side of biomedical publication you surely know very well the Weinberg-Spandidos-Siminovitch story from the dawn of oncogene research. A retrospective soundbite here for those too young to have been informed

        More detail is available for the curious in the book by Natalie Angier

        Cripes! What chance would a naive postgrad ever have in such an environment?

        1. In reply to Scrutineer April 22, 2014 at 1:23 pm

          Another paper, J Immunol. 2008 May 1;180(9):6199-210, with the company of DA Spandidos, publisher of multiple journals.

          Highly informative critique at Pubpeer:

          The imgur image of figure 3 is worth a look.

          For reference:

          J Immunol. 2008 May 1;180(9):6199-210.
          Inhibition of Yin Yang 1-dependent repressor activity of DR5 transcription and expression by the novel proteasome inhibitor NPI-0052 contributes to its TRAIL-enhanced apoptosis in cancer cells.

          Stavroula Baritaki2,*, Eriko Suzuki2,*,‡, Kazuo Umezawa‡, Demetrios A. Spandidos§,
          James Berenson¶, Tracy R. Daniels†, Manuel L. Penichet*,†, Ali R. Jazirehi*,
          Michael Palladino∥ and Benjamin Bonavida3,*

          – Author Affiliations

          *Department of Microbiology, Immunology, and Molecular Genetics and
          †Department of Surgery, Division of Surgical Oncology, David Geffen School of Medicine, Johnson Comprehensive Cancer Center, University of California, Los Angeles, CA 90095;
          ‡Department of Applied Chemistry, Keio University, Yokohama, Japan;
          §Department of Clinical Virology, Faculty of Medicine, University of Crete, Heraklion, Greece;
          ¶Institute for Myeloma and Bone Cancer Research, West Hollywood, CA 90069; and
          ∥Nereus Pharmaceuticals, Inc., San Diego, CA 92121

          PMID: 18424742

          1. Yes that is one mightily impressive figure. Must have required a great team effort!

            Spandidos recently featured here in the comments of an article on Weinberg –


            I know that you will remember it well, David.

            Checking PubMed, Spandidos first began to publish in the mid-70s, Bonavida somewhat earlier in the late 60s. I was especially surprised by the longevity of the latter. Despite Bonavida’s fecund publication record (386 PubMed entries), Spandidos comfortably bests him (476 PubMed entries), clearly not seriously hampered by the “Case of the unpurchased Petri dishes” alluded to above.

            It is rather fascinating how these two venerable scientific contributors have come into a sharp PPPR focus at the tail end of their careers when they should be looking forward to well-earned rest.

            Around about the time they started publishing, wasn’t there a legendary song by The Who that encapsulated the Zeitgeist? Trying to remember the line that comes after this one

            “Then I’ll get on my knees and pray

      4. Continuation david hardman April 21, 2014 at 2:27 am

        Some difficult to explain data in Mol Cancer Ther. 2004 Jan;3(1):71-84 Mol Cancer Ther. 2003 Nov;2(11):1183-93 discussed here:

        For reference:

        1. Mol Cancer Ther. 2004 Jan;3(1):71-84.
        Resveratrol modifies the expression of apoptotic regulatory proteins and sensitizes non-Hodgkin’s lymphoma and multiple myeloma cell lines to paclitaxel-induced apoptosis.
        Jazirehi AR1, Bonavida B.
        Author information

        1Department of Microbiology, Immunology and Molecular Genetics and Jonsson Comprehensive Cancer Center, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA 90095, USA.

        2. Mol Cancer Ther. 2003 Nov;2(11):1183-93.
        Rituximab (anti-CD20) selectively modifies Bcl-xL and apoptosis protease activating factor-1 (Apaf-1) expression and sensitizes human non-Hodgkin’s lymphoma B cell lines to paclitaxel-induced apoptosis.
        Jazirehi AR1, Gan XH, De Vos S, Emmanouilides C, Bonavida B.
        Author information

        1Department of Microbiology, Immunology and Molecular Genetics, David Geffen School of Medicine at UCLA, University of California, Los Angeles, CA 90095, USA.Mol Cancer Ther. 2003 Nov;2(11):1183-93.

    2. The speed with which she has asked for a retraction might indicate that she has long been unhappy about this work. This would further imply that it is not in her nature to be a ready fabricator.

      I might be reading too much into this, but she has not apologised to her coauthors, only to the journal. Time will probably tell whether that omission is significant.

      1. Indeed – and in her most recent comment, she says “I declare that I am not in charge regarding inappropriate corrections in Int. J. Oncol. 2010. The paper was prepared by Dr. Bonavida group [her former lab] after my study in UCLA.”

        This paper (of which Suzuki is first author) had been highlighted in one comment as another case of possible image manipulation.

        So in her comment, Suzuki seems to be saying: I take responsibility for one paper, but not the rest of them… in which case the Bonavida group might have some explaining to do.

    3. AF, can you point these out please? If this work was done when Suzuki was a PhD student, the responsibility falls with her supervisor(s) as it is they, as University representatives, who ‘own’ the data, not Suzuki.

      Did no-one check the ORIGINAL DATA?

      Suzuki should be commended for honesty, but she is not solely to blame.

      Surely the senior laboratory members (where are they?) should comment. Afterall, they are willing to take the glory and promotions when papers are published and used to support grant applications, so why don’t they take responsibility? If they are senior academics, what does this say of senior academics?

      Were any of Suzukis papers/research data used to support any grant applications?

  3. If Keio have any sense they will look at the entire thesis and judge the doughnut, and not the hole (to mix my metaphors) and then come to decision as to the ratio of doughnut to hole.

    1. Given the high tension in Japan these days over STAPs, the hole is probably all that they are looking at.

    2. i completely agree. problem is now in japan the science institutions want to take plenty of time and investigate carefully, but media and public opinion are putting extreme pressure on them to make a judgment, quickly and severely.

  4. I do not know the authors, but I was told it would be extremely difficult for women in Japan to succeed in a scientific and usually men-driven career in Japan. I post this only as a hypothesis – I have no statistics of female professors in Japan to support the hypothesis. That of course can not excuse all mistakes, but may explain the pressure for some who may want to live in Japan.

    1. please use data or at least first-hand anecdotes. women are rare in Japanese science, but that doesn’t mean they are persecuted. ms. obokata (of STAP fame) was very successful at an extremely young age, some would say more successful than a typical man, perhaps because she always gets noticed.

  5. The punishment (although it hasn’t been yet meted out) certainly does not fit the crime if this is the only thing the author will be accused of, and this has been my observation with misconduct cases in Japan. Maybe it’s a cultural thing, where the boss has to take the fall like a disgraced samurai, whereas in the US one paper with mistakes does not kill a career.

    But it also seems to me that women get harsher punishments. There was a case of a famous chemist who put her grant money in a savings account while it was sitting around and said that as long as it’s there, it can make some percentage for the group. She was fired from the university despite having a very successful and large research group. I’m not sure whether a man would have had the same level of punishment.

    Plus there are these big shows of punishing the person responsible for every small infraction, yet in Japan very real criminals get away with creating parallel societies outside the law (Yakuza) and faking reports (Fukushima power plant safety review) without many consequences. Obviously women are rare in politics, business CEOs, and Yakuza bosses, so I don’t agree with punishing a person very harshly for an offense that in the US wouldn’t result in this kind of censure. It results in a discriminatory double standard and it’s hypocritical if the rest of society isn’t cleaned up.

    If you haven’t guessed it yet, I’m very much against revoking her PhD if this is the only incident and it’s proven to be an honest mistake as she claims.

    1. there is no evidence of women getting harsher punishments than men, and suzuki has not received any punishment yet anyway. plus, suzuki was not the “boss”, so she doesn’t fit into the theory of the boss taking the fall for some minor infraction. and why does rarity of women leaders affect the weight of punishment suzuki should receive?

    1. It is totally unrealistic to think there is any chance of Suzuki staying in science, regardless of the pressures she may have been under at the time. If she kept a diary, she would be able to write a book of her experiences – I would buy it. Perhaps such a book could even contribute toward an overhaul of the US research system?

      Given her speed of response and the associated remorsefulness, we may follow Shakespeare to hope that she receives sympathy and understanding in our expectation that

      “Nothing in her scientific life will become her like the leaving of it.”

      1. Scrutineer

        Why shouldn’t Suzuki stay in science?

        There are far worse science-fraudsters who remain entrenched in their positions, employing technicians, postdocs, advising students…….they are far more dangerous than Suzuki will ever be.

        As Bill also wrote “suspicion always haunts the guilty mind”

        1. The handicap would always be there. Write a grant – referees can’t be expected to ignore the past. Apply for promotion – the safe option for the committee is to appoint someone else. My personal opinion is that it can’t work.

          People who have the bad luck to start their scientific careers in a corrupted lab environment have it stacked against them, regardless of how honest they may be personally. The case of the honest postgrads and their corrupt PI is relevant here.

          There may well be “far worse science-fraudsters who remain entrenched in their positions” but I presume it is difficult to find one that was not steadfast in denial of any wrong doings. Anyway, PPPR is hopefully going to provide the weight of evidence to flush out an entrenched miscreant or too…

          1. Exactly. This is an issue in my country, where fraudster scientists will favour other misdoers around them as to keep the environment favourable to their image. I guess this happens at some places in most parts of the world. People with justly tarnished records can always find their place in such places. It is like the Slytherins, and maybe like in fiction they also play their role in the big picture.

  6. When young, first authors take all blame and go public immediately on foul play, one should always suspect of some greater foul play. I have seen more than one case where the one taking all blame has been offered protection and compensation to do so, and after few years is actually not in a bad economical/political situation at all. Bad guys always exploit the naive kindness of good guys.

  7. Agree that her response is disproportionate, unless there is more (such as, it was deliberate and she was a knowing participant). I don’t want to fall back on stereotypes, but there seems to be a “fall on your sword” mentality at work here. Also agree that there seems to be much more to the story at the head PIs lab.

    1. “fall on your sword” only has meaning if you are saving someone else, and there is no evidence of that in this case. it seems to me (here in japan) that ms. suzuki wants to show she’s different from ms. “stap” obokata. ms suzuki may feel like female scientists have taken a black eye, because the media here continue to make a HUGE deal about obokata’s gender (and age, and and attractiveness, and history of working at Harvard).

  8. “I just realized that I had misconducted pasting some JPG images as I had huge amounts of similar data at that time.” Having huge amounts of similar data is all the more reason to double- and triple-check things, and having someone else verify that you have done things correctly. Kind of hard to believe that this was an honest mistake. Nonetheless, Dr. Suzuki deserves credit for coming clean. Retroactive honesty is better than no honesty at all.

    1. Agreed Freeheeler, and as we are on Shakespeare “No legacy is as rich as honesty”

      If Suzuki can do it, many can.

  9. 1. The huge amount of papers (including many at good journals) with image manipulation is troubling. If this kind of pretty obvious and smoking gun stuff is going on, what else is happening on parts of the text or with images that are false (say from wrong species or the like) but that aren’t duplications. I think we may have a much wider problem.

    2. Impression I get is that the problem is more in medical/biology than in chemistry and physics. I’m not sure why this is the case, but it almost seems like having more money (health sciences get huge funding compared to physical sciences) leads to more dishonesty not less. Perhaps it attracts more people into the field and you get a different dynamics with more people, even more low performers?

    1. In reply to exchem April 21, 2014 at 1:22 pm

      “Impression I get is that the problem is more in medical/biology than in chemistry and physics.”

      I think you are probably correct. This has been discussed before (no claim of originality on my part), but it is good to discuss it again. Physics and chemistry have much more hard data which has been accepted for longer.

      Recently a bright spark in India noticed that something was amiss.

      “This claim that the authors were able to dissolve the non-polar pentacyclic triterpenoid lantadene A in water was quite surprising to me, as the Toxnet database reports the calculated solubility of lantadene A in water at 25 °C as just 7.68 × 10−5 mg/L [2], suggesting that lantadene A is essentially insoluble in water. The facile solubilization of lantadene A in water mentioned in this paper thus raises questions about the identity of the compound actually tested and therefore the whole study and its conclusions”

      In biology there may be fewer simple, hard facts. Given these hard facts anybody who has done high school physics or chemistry is unlikely to argue that the findings are correct, or at least their superiours will be less likely to back them up. It might also be that people in medical/biology fields have forgotten their high-school physics and chemistry.

    2. Re: your point No.2: You may be right. I regard the ‘medical/biology’ error rate as being higher because there is more pressure at that time to avoid or bury “bad news”. For example, if a pharma product is at its final stage before licensing and release then discovering errors at that point could possibly blow millions of dollars of investment. By comparison, if one messes up at the earlier ‘chemistry/physics’ stage, one is far less likely to have both the Sales and Marketing departments chasing you down the hallway and more likely to be able to do a snow-job and re-test at far lower expense.

      1. I think the issue with med/bio is simpler than that — faked western blots are among the easiest results to catch, therefore we catch a lot of them. Data that can be faked without pictures, entirely in a spreadsheet or other computer file, are harder to catch.

        1. In reply to StrongDreams April 21, 2014 at 1:54 pm

          “faked western blots are among the easiest results to catch, therefore we catch a lot of them”.

          Yes, the western blot is a good fake detection assay.

          “Data that can be faked without pictures, entirely in a spreadsheet or other computer file” are part of the abyss.

          1. I think faked crystallographic data is very rare (never heard of it). People make mistakes on structure identification or have a different substance than they think they do sometimes. But those are all honest mistakes, not fabrication. In any case the data is archived (routinely, prior to publication) and reviewers or readers can resolve the structure, look for mistakes, etc.

  10. Eriko Suzuki second author on 2016 retraction of 2008 J Immunol paper.

    J Immunol. 2008 May 1;180(9):6199-210.
    Inhibition of Yin Yang 1-dependent repressor activity of DR5 transcription and expression by the novel proteasome inhibitor NPI-0052 contributes to its TRAIL-enhanced apoptosis in cancer cells.
    Baritaki S1, Suzuki E, Umezawa K, Spandidos DA, Berenson J, Daniels TR, Penichet ML, Jazirehi AR, Palladino M, Bonavida B.
    Author information
    1Department of Microbiology, Immunology, and Molecular Genetics, Division of Surgical Oncology, David Geffen School of Medicine, University of California-Los Angeles, 10833 Le Conte Avenue, Los Angeles, CA 90095, USA.

    2016 retraction notice.

    Pubpeer comments.

      1. Retraction: Inhibition of Yin Yang 1-Dependent Repressor Activity of DR5 Transcription and Expression by the Novel Proteasome Inhibitor NPI-0052 Contributes to its TRAIL-Enhanced Apoptosis in Cancer Cells
        A request was received from the Acting Research Integrity Officer of the University of California, Los Angeles (UCLA), to retract this article: “Inhibition of Yin Yang 1-Dependent Repressor Activity of DR5 Transcription and Expression by the Novel Proteasome Inhibitor NPI-0052 Contributes to its TRAIL-Enhanced Apoptosis in Cancer Cells” by Stavroula Baritaki, Eriko Suzuki, Kazuo Umezawa, Demetrios A. Spandidos, James Berenson, Tracy R. Daniels, Manuel L. Penichet, Ali R. Jazirehi, Michael Palladino, and Benjamin Bonavida, The Journal of Immunology, 2008, 180: 6199–6210.

        The Editor-in-Chief of The Journal of Immunology was informed that UCLA had conducted a review of the work and concluded that data used in some of the figures in this article could not be supported. In particular, issues involving duplication of images were identified in Figs. 2A, 3B, and 6B. The article is therefore retracted.

        It is noted that not all coauthors of this article worked on the data presented in the figures identified by UCLA. In particular, Manuel L. Penichet has advised the Editor-in-Chief that his contributions and those of Tracy R. Daniels were confined to data in Figure 7.

        Copyright © 2016 by The American Association of Immunologists, Inc.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.