Surgery chair who blamed image issues on software logs three more retractions

A researcher who claimed image problems in a retracted paper were the result of a software glitch, and not intentional, has lost three more papers — all for image manipulation.

In two notices, the Journal of Biological Chemistry specifies that duplicated images were used to represent different experimental conditions; one notice simply says the paper was affected by image manipulation.

All of the notices specify the papers are being retracted by the publisher, the American Society for Biochemistry and Molecular Biology — which this month published a set of recommendations for preparing a paper, including how to avoid excessive manipulation.

The papers were published between 2002 and 2010, and all share the same last author (Paul Kuo, currently chair of surgery at Loyola Medicine) and first author (Hongtao Guo, at Duke).

Here’s the first notice:

This article has been retracted by the publisher. The same data were reused to represent different experimental conditions. Specifically, lanes 2–4 and 6–8 of the actin immunoblot in Fig. 3A were duplicated. Additionally, the nuclear P-STAT1 immunoblot from Fig. 3A was reused as P-STAT1 in Fig. 3B.

Osteopontin and protein kinase C regulate PDLIM2 activation and STAT1 ubiquitination in LPS-treated murine macrophages” was published in 2010, and has accumulated eight citations, according to Clarivate Analytics’ Web of Science, formerly part of Thomson Reuters. Some of the images in the paper have been discussed on PubPeer.

Here’s the second notice, for “Hepatocyte nuclear factor-4α mediates redox sensitivity of inducible nitric-oxide synthase gene transcription:”

This article has been retracted by the publisher. The same data were reused to represent different experimental conditions. Specifically, in Fig. 1, lane 3 was reused in lane 11, and lanes 9 and 10 were reused in 17 and 18. Additionally, lane 3 from Fig. 1 was reused in lane 2 from Fig. 3. Finally, lanes 2–4 from Fig. 1 were flipped horizontally and reused in lanes 5–7 from the left panel of Fig. 4.

The 2002 paper has been cited 16 times; it has also been discussed on PubPeer.

Finally, here’s the notice for “Characterization of short range DNA looping in endotoxin-mediated transcription of the murine inducible nitric-oxide synthase (iNOS) gene:”

This article has been retracted by the publisher. Fig. 2, B and C, and Fig. 4B were inappropriately manipulated.

The 2008 paper has been cited 15 times; here’s its PubPeer entry.

Kuo — who was at Duke at the time the research was conducted — has two previous retractions; one, retracted in 2014, included Guo as first author. The journal said it was retracting the paper because one of the images contained a black box; Kuo later told us the black box was the result of a glitch in the software used to prepare the manuscript, not intentional manipulation.

Kaoru Sakabe, data integrity manager for the American Society for Biochemistry and Molecular Biology, this month offered some helpful advice to authors who want to avoid image problems (unrelated to the aforementioned retractions). As she writes in a Perspectives article:

I’ve heard a lot of different excuses from authors I’ve investigated for violations, such as erasing blemishes and bands, reusing data from different publications and cutting and pasting bands to create data that never existed. These excuses run the gamut from somewhat credible to incredible — although I haven’t yet heard that someone’s dog ate it…in brief, you should manipulate your image as little as possible when preparing the figures for publication. Your final image should be a true representation of the film or image when you captured the original.

We contacted Sakabe, who told us the timing of her article and Kuo’s retractions was purely coincidental:

The Due Diligence column is a series that was in development for several months before the first article was published and will continue for a while.  It is part of our larger program to educate the community to avoid these issues.

She explained why the retractions say they were prompted by the publisher:

Retractions are initiated by the publisher and approved by the ASBMB Publications Committee when authors do not agree to withdraw their papers.

We contacted Guo. When we emailed Kuo, we received an automatic reply saying he was away with limited email access.

Update 1/26/17 11:23 a.m. eastern: We’ve heard from Kuo, who told us:

Our research group is disappointed in the JBC retractions. We have gone to extreme measures to address the JBC queries. However, our experience has been extremely frustrating.

1. JBC lacks an SOP for addressing these issues. If they do have one, they will not share it. There is no ombudsman. The existing methodology is extremely opaque and inconsistent with current industry standards as set forth by COPE and recent publications in Cell Press and Plant Physiology. Overall, their approach borders on absurd in a fashion reminiscent of Kafka’s novel, The Trial.

2. At the time of submission of these papers, JBC did not have data retention or image presentation guidelines. Current guidelines were established after 2010. The intervening time period for the 3 papers exceeds the standards for the time, which we argue is the NIH standard. Selective retrospective application of current JBC image standards to our papers while ignoring the same data presentation technique in multiple papers from the same volume is irrational. (And we submitted the references of these JBC papers.) Asking the lab to address queries re: similar appearing bands in DNA loading wells when the corresponding data sections are different is equally bizarre.

3. The JBC decision to ignore the exculpatory conclusion of the official Duke institutional review indicating no research misconduct and no misrepresentation of the scientific record, corroborating evidence from contemporaneous experimental replicates, published independent confirmation of our findings by other labs, and confirmatory results from our recent experiments performed using the original samples is equally mystifying.

4. Of even greater concern was the superficial and often flawed analysis of the images in the publications and additional submitted data. For example, in one instance where it is claimed that same data were used to represent different conditions, a simple perusal of the figure legends shows that the conditions were the SAME. When we pointed this out, it was ignored.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

21 thoughts on “Surgery chair who blamed image issues on software logs three more retractions”

  1. Immunol Lett. 2002 Jan 1;80(1):21-6.
    Nitric oxide is necessary for CC-class chemokine expression in endotoxin-stimulated ANA-1 murine macrophages.
    Guo HT1, Cai CQ, Schroeder RA, Kuo PC.
    Author information
    1Department of Surgery, Duke University Medical Center, 102 Bell Building, Box 3522, Durham, NC 27710, USA.

    Figure 3.

    Figure 4.

    Please compare figure 3 Immunol Lett. 2002 Jan 1;80(1):21-6 with figure 4 J Immunol. 2001 Jan 15;166(2):1079-86.

    For reference. J Immunol 166:1079.

  2. J Immunol. 2001 Jan 15;166(2):1079-86.
    Osteopontin is a negative feedback regulator of nitric oxide synthesis in murine macrophages.
    Guo H1, Cai CQ, Schroeder RA, Kuo PC.
    Author information
    1Department of Surgery, Georgetown University Hospital, Washington, D. C. 20007, USA.

    Figure 3.

  3. The embedded link (above) to the ASBMBToday web advice by Kaoru Sakabe (on avoiding image manipulation) revealed a great idea, I hope ASBMB does follow through, by adding others of the same topic; and hopefully someday collecting all together as a useful repository for students, authors and art editors alike. It illustrates that one of the simple means by which journals and authors alike can protect their ‘product’ from the vagaries of post-publication peer review: Namely, to require any image data that they publish contain sufficient detail, features, and background that make the results identifiably unique (as a representation of data). Ideally, the published image should contain sufficient detail to establish its own credibility as original data, and thus insulate it from challenge to its authenticity.

      1. Because both Author and Journal have an investment in the effort going into publication. Realizing this as an economic matter should encourage the journals to ‘up’ the print quality of what they publish. My reasoning here is that when the image is printed with sufficient detail, then the effort they have to put into resolving what to do with an allegation is greatly reduced. Then a questioned image just doesn’t wag its tail, it barks for attention. Quality serves both interests: It provides a tangible means to ask, should I take this anonymous concern seriously? It makes decision about disposition easier, and it makes it harder to avoid doing the right thing. Conversely, with a good image the merits of any allegation can also be better evaluated. (After all, if we are scientist, we ought to be able to form judgments about data.) Very little has been said about the role that poor print quality in journal images plays in the downstream grief associated with dealing with allegations in post-publication peer review.

        1. So where do you think a retraction has happened because of low quality images?
          Do you think that the images issues pointed out are equivalent to typos?

          “role that poor print quality in journal images plays in the downstream grief associated with dealing with allegations in post-publication peer review.” If the auhtors were happy enough with the quality of the images why didn’t they complain to the journals soon after publication?

          1. Should be “if authors were unhappy with quality of images why didn’t they complain”.

            “require any image data that they publish contain sufficient detail”.
            Reasonable enough, but it will not work without people checking within and between publications.

            It does also leave open the question of what is sufficient data. In any events judements will not be according to image quality alone. For example people will still have to judge if the similarity between images (which are not identical) are significant, or not. This may include at looking at several publications by the authors under questions and also at some by some comparator authors. In short people need to use their nous and expend time and effort rather than rely on panaceas.

          2. The point being missed here are the retractions that *did not* occur because the challenged data were too lousy, too ambiguous to be effectively tested! And how much easier it would be for journals and readers to decide the merits of anonymous allegations concerning image data that are otherwise too ambiguous to test, or for that matter, to trust? Faced with decisions on whether to correct the literature, are journals now, essentially, paying the overhead for publishing poor quality image data? Small steps would go a long way towards improving the integrity of the literature.

  4. Dig Dis Sci. 2005 Jul;50(7):1288-98.
    Nitric oxide-dependent osteopontin expression induces metastatic behavior in HepG2 cells.
    Guo H1, Marroquin CE, Wai PY, Kuo PC.
    Author information

    1Department of Surgery, Duke University Medical Center, Durham, North Carolina 27710, USA.

    Figure 1.

  5. Carcinogenesis. 2006 Jun;27(6):1134-45. Epub 2006 Feb 12.
    Integrin-linked kinase regulates osteopontin-dependent MMP-2 and uPA expression to convey metastatic function in murine mammary epithelial cancer cells.
    Mi Z1, Guo H, Wai PY, Gao C, Kuo PC.
    Author information
    1Department of Surgery, Duke University Medical Center, Durham, NC 27710, USA.

    Figure 1B.

    Figure 2.

    Figure 3A.

  6. 2017 correction figures 2 and 8 J Biol Chem. 2004 Mar 19;279(12):11236-43.

    J Biol Chem. 2004 Mar 19;279(12):11236-43. Epub 2004 Jan 13.
    S-nitrosylation of heterogeneous nuclear ribonucleoprotein A/B regulates osteopontin transcription in endotoxin-stimulated murine macrophages.
    Gao C1, Guo H, Wei J, Mi Z, Wai P, Kuo PC.
    Author information
    1Department of Surgery, Duke University Medical Center, Durham, North Carolina 27710, USA.

  7. Re-using an image is a problem with working ethic which is a matter of trustworthiness; whether a data is replicated and whether a conclusion is changed or not are the matters of science. They are completely different.

Leave a Reply

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