Retraction Watch

Tracking retractions as a window into the scientific process

Third retraction for former rising star found guilty of misconduct

with 7 comments

A once-prominent researcher in the field of infectious disease — who was found guilty of misconduct last year— has had a third paper retracted, a 2006 article in PNAS.

Last year, the University of Dundee in Scotland found that Robert Ryan had committed research misconduct, which included misrepresenting clinical data and duplicating images in a dozen different publications. After a failed attempt to appeal the decision, Ryan resigned.

In April, we covered Ryan’s first two retractions – a 2012 paper in Molecular Microbiology, which cited image errors, and a 2011 paper in Journal of Bacteriology, which cited image duplication.

Now, PNAS has retracted a 2006 paper, which cites potential image duplication as well as “irregularities” in the data.

Here’s the retraction notice for “Cell–cell signaling in Xanthomonas campestris involves an HD-GYP domain protein that functions in cyclic di-GMP turnover:”

Retraction for “Cell–cell signaling in Xanthomonas campestris involves an HD-GYP domain protein that functions in cyclic di-GMP turnover,” by Robert P. Ryan, Yvonne Fouhy, Jean F. Lucey, Lisa C. Crossman, Stephen Spiro, Ya-Wen He, Lian-Hui Zhang, Stephan Heeb, Miguel Cámara, Paul Williams, and J. Maxwell Dow, which was first published April 12, 2006; 10.1073/pnas.0600345103 (Proc Natl Acad Sci USA 103:6712–6717). The authors wish to note the following: “After publication it was brought to our attention that Fig. 3 A and B appear to be identical, although they show different enzyme activities, and likewise Fig. 4A is very similar to Fig. 4B. In addition, there were irregularities in the Western blot shown in Fig. 4C. Unfortunately, we cannot locate the complete raw data used to construct these figures, although repeated experiments using the original strains support the findings. Based upon the number of issues, we hereby retract the work. We apologize to our colleagues and the scientific community for any inconvenience this might have caused.”

The paper, on which Ryan is first author, has been cited 340 times, according to Clarivate Analytics’ Web of Science.

Ryan told us:

I believe the retraction statement and the posts on Pubpeer explain everything. … When we discovered the issues I did write to the journal directly to highlight the errors before they appeared on Pubpeer. I was not involved in generating the data that was misrepresented but like the other authors must take responsibility for missing these issues in the generation of the manuscript figures. I am terribly sorry about that.

Last year, commenters on PubPeer flagged multiple instances of possible image duplication in the paper.

Ryan added:

It should be stressed that independent labs to date have repeated all of this work and shown similar outcomes. It is my understanding that given there were errors in several figures the authors wished to retract and republish the article rather than correct it. Like the other articles that have mistakes and have had to be retracted, I have provided the strains and information to independent scientists to republish. I will provide links to the preprint and articles once they are accepted.

Ryan has won several prestigious awards for his research on infections that can be deadly in people with lung diseases, including the prestigious Society for General Microbiology’s Fleming Prize (which the society withdrew in March 2017). In 2015, he was also named an EMBO Young Investigator, but the program withdrew his name in 2016.

Ryan has five corrections, three of which we’ve covered here and here. The two most recent corrections, both published in Molecular Microbiology this past April, also flag image duplication but point the finger at other researchers.

Here’s the correction notice for the 2013 paper in which Ji-Liang Tang—who is based at Guangxi University in China—“takes full responsibility” for the original error:

The authors have been made aware of issues with Fig. 3A of this paper which illustrates the different classes of virulence as determined by lesion length. In constructing this figure, leaf images reported in two other publications (An et al., 2013; EMBO J 32(18): 2430–2438 and O’Connell et al. 2013; Mol Plant Microbe Interact 26(10): 1131–1137) were used to represent Class 0 and Class IV symptoms, respectively. To address this duplication, we have replaced these panels in Fig. 3A. Professor J. L. Tang takes full responsibility for the original error. Together he and the corresponding author take accountability for the authenticity of the correction. All authors sincerely apologize for the error and emphasize that the quantitative data that is presented is correct and fully supports the conclusions drawn.

The paper, “High-resolution transcriptional analysis of the regulatory influence of cell-to-cell signalling reveals novel genes that contribute to Xanthomonas phytopathogenesis,” on which Ryan is corresponding author and Tang is a middle author, has been cited 31 times.

And here’s the corrigendum for the 2006 paper, “Cyclic di-GMP signalling in the virulence and environmental adaptation of Xanthomonas campestris,” in which both Tang and Maxwell Dow at University College Cork take “full responsibility:”

The authors have been made aware of issues with Fig. 1 and Supplementary Fig. S4 of this paper. In Fig. 1, the same leaf image labelled ‘ddH2O control’ was used both in this paper and in McCarthy et al. (2008) Mol Plant Pathol 9(6): 819–824 to represent the negative (un-inoculated) control. We believe this duplication occurred as the data sets for these two papers were generated from assays carried out at the same time. We have replaced this panel of Fig. 1to address this issue of image duplication between the two articles. In Supplementary Fig. S4, errors in construction have led to duplication of images between columns 1 and 3, although the findings as described within the text of the paper are correct. In order to resolve this issue, we have repeated this experiment (with essentially the same outcome) and constructed a replacement Figure. The original quantitative data taken at the time of these experiments was provided to the journal for review. None of the conclusions of the paper are altered. This correction has been submitted by J. L. Tang and both corresponding authors (J. L. T. and J. M. D.) take full responsibility for the original errors and authenticity of the correction. We sincerely apologise to the readers for the errors and appreciate the opportunity to correct the scientific record.

The paper, on which Ryan is first author and Dow and Tang are corresponding authors, has been cited 93 times.

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.

  • JThomas August 8, 2017 at 8:56 am

    How many of these people are thought of as “rising stars” before the bottom falls out…
    I even had a colleague refer to one as a “miracle worker” before the trouble started.

  • fernandopessoa August 8, 2017 at 10:02 am

    the universities fall for it all the time

  • Scott Allen August 8, 2017 at 11:36 am

    These three papers were cited over 450 times and were used for the basis or in support of other papers that quoted the cites. Since the original building blocks for those other papers have been shown to be invalid, how many of those other papers will be retracted ?
    It is just like putting a defective part in a product, removing the defective part does not make the product, in fact it reduces the reliability of the whole product.

  • JThomas August 8, 2017 at 2:44 pm

    Scott Allen has a great point. Either subsequent citing papers suffer from citing fakery, or we’re all playing a game. More directly, if the work we cite is fake and that doesn’t impact our work, proposals, conclusions, then we’re really not basing anything on the work we cite. Then what does the whole scholarship construct even mean?

    I’ve seen it increasingly on RW; more and more commenters seem to be suggesting it: make academic/scientific fraudsters face civil and/or criminal penalties. I agree with others that this is the only way to curtail these phonies. Otherwise, at some point the entire enterprise collapses around its rotten foundation.

  • PJTV August 8, 2017 at 3:44 pm

    The concern of Scott Allen is real, we are all building on the work of others and if that work proved to be the result of misconduct, we will all suffer. In this case 450 articles were affected. Obviously, the rate of retractions is minuscule and their effect minimal on a large scale. However, the conditions creating misconduct (publish or perish, fight for funding, etc) are also worsening the situation. Also, scientists are under scrutiny in a world where fake information became common.
    What happens so far is that work citing retracted work is never or hardly questioned. Maybe scientists should take this more seriously, for example by investigating the cited works whether those should not be amended this in order to restore trust in science as a whole. But who should take here the responsibility? Individual scientists, institutions, journals, learned societies?

  • rfg August 8, 2017 at 6:13 pm

    Civil and criminal penalties for persons that are proven to have committed misconduct is appropriate and may have some deterrent affect.

    What is needed is that institutions should be required to payback the fraudulently obtained grant or contract funds. Multiplier penalties should be imposed on the institution for obstructing or delaying the investigation.

  • JThomas August 9, 2017 at 8:44 am

    rfg – I guess that your opinion is fueled by frustration that misconduct seems to go unabated with nobody facing any real risk of significant penalties. If I’m correct, I feel that frustration, too. However, I think we must also be careful how we attempt to solve the problem.

    There may be some situations in which the university is culpable to some extent. As one who has conducted misconduct investigations, I can write with some authority that determining the extent of a host institution’s culpability would require investigation. It may also decrease, even further, an institution’s willingness to cooperate, “come clean” and not sweep its dirty secrets under the rug.

    A different issue, for me at least, is my belief that the guilty – and ONLY the guilty – should face penalties. Legal people might argue at some point what the definition of “guilty” is. But for now, I think a first step is to hold those who obtain public research funds by fraudulent means to the same punishment standards and jeopardy that any fraudster must face. When this has happened (and I believe it will eventually) it might be a more appropriate time to discuss systemic issues that might unintentionally promote fraud.

  • Post a comment

    Threaded commenting powered by interconnect/it code.