The domino effect: More retractions of papers that cited retracted PLoS ONE GMO cassava study

Last month, we reported on a PLoS ONE paper about genetically modified cassava — or, more correctly, allegedly GMO cassava — that was being retracted because data “could not be found.” We have an update on that story, namely that a paper relying on materials from that lab will be retracted, and that authors of a review that included a figure from the graduate student who claimed to have done the work will retract part of their paper.

As a Retraction Watch commenter on our earlier post noted, referring to Claude Fauquet, the PI of the Danforth Center lab where graduate student Mohammad Abhary worked:

The study (Biotechnologically-Modified Cassava: Protein Absorption Relative to Casein, Li. et al., 2012) used cassava obtained from Mr. Fauquet to show that zeolin-enhanced cassava allowed mice to grow, and non-enhanced cassava resulting in weight loss and lack of growth in the control mice group. Do the findings and methodology of this study need to be checked too?!

The link to the study now returns a “Service Unavailable” page. The commenter, as is often the case, was onto something. We asked the senior author of that study, Kendal Hirschi of Baylor College of Medicine in Houston, for an update. He told us early last week:

The retraction  is in the works- the journal requires a signed document from all four authors and some of the co-authors are now in China.

We informed the journal immediately about our plans to retract the article and have worked on this since the minute (mid-september) we found out about the samples being tainted.

As far as we can tell, Hirschi’s lab was a victim here, having been sent materials by Fauquet — also by all accounts unaware of the problem with the cassava — and now they’re trying to clean up the scientific record. A retraction is never an easy thing to swallow, particularly when the error wasn’t even your own fault, but the authors should be commended for swift and appropriate action.

We’ve learned that the other retraction is of part of a paper in the Annual Review of Plant Biology, 2011’s “The BioCassava Plus Program: Biofortification of Cassava for Sub-Saharan Africa.” That paper includes authors from a number of institutions, including first author Richard Sayre of the Danforth Center, and Fauquet. The authors are retracting Figure 3 of the review, which is based on previously unpublished data from Abhary that can’t be confirmed, and a paragraph on page 259 that describes the PLoS ONE paper.

The review has been cited 10 times, according to Thomson Scientific’s Web of Knowledge. We tried reaching Sayre for comment, and will update with anything we learn.

15 thoughts on “The domino effect: More retractions of papers that cited retracted PLoS ONE GMO cassava study”

  1. Originally submitted on 2014/09/25 at 8:17 am; reposted after site migration

    I was stunned to learn of this story only yesterday. I studied tissue culture and genetic transformation of cassava as an undergraduate student using 12 African cultivars, so that experience catapulted me, in theory and effectively, into plant science. The researchers from the Donald Danforth Research Center are role models, and leaders, especially in cassava biotechnology, so this retraction of the 2011 PLOS ONE paper sent a ripple effect throughout the cassava research field globally. Given the fact that so many high level scholars are embroiled in this saga, given the fact that the Donald Danforth Center was, and continues, to receive funding from the Bill and Melinda Gates Foundation, and given the fact that all suddenly went very, very silent from about September/October 2012 onwards, I decided to explore this topic and case further. I can now with confidence claim, after having spent the better part of today examining the details, state that this is most likely an even higher profile case, given its potential economic and political fall-out, than the multiple Vivanco [1] and Ben Scheres [2] retraction cases. I hope, in the next few days, to provide updates, critiques and suggestions, even a hypothesis, about why this case suddenly went silent and why the plant science community has not been outraged, why justice has not been duly served and why the Donald Danforth Research center continues to operate without any consequences. Did I already state that the Donald Danforth Research center is one of plant science’s premier research centers on the globe? And did I already state how deeply disappointed I am in these three role models (Vivanco, Scheres and Taylor/Fauquet and associated researchers), who formed an integral part of my undergraduate and graduate inspiration? Personally, these cases are making me feel very sick, literally physically sick, to the depth of my stomache.

  2. Submitted on 2014/09/25 at 10:20 pm | In reply to JATdS; reposted after site migration

    I wish to provide the first update and commentary which serves to set the record straight about this two-year-old case. An update is important since several of the links provided in the other RW web-page related to this story are not working any longer. This highlights how important it is to document web content through screen-shots and self-archived documents since authorities tend to remove such documents to avoid negative publicity, or their tracking. The first update relates to the down-stream effect of the original false PLOS ONE GM cassava paper, as indicated by the title of the story “The domino effect: More retractions of papers that cited retracted PLoS ONE GMO cassava study”. RW stated that “We have an update on that story, namely that a paper relying on materials from that lab will be retracted, and that authors of a review that included a figure from the graduate student who claimed to have done the work will retract part of their paper.” Kendal Hirschi also stated that “The retraction is in the works”, when referring to an OMICS published paper [1]. So, why has this supposed-to-be-retracted paper that simply required the signature of the co-authors, not been retracted?

    Separately, an Annual Reviews in Plant Biology paper, which was subjected only to an erratum, also included Richard Sayre, another highly respected plant scientist, at the Donald Danforth Plant Science Center (DDPSC), but who is no longer listed there [2]. The RW story was published on October 9, 2012. The erratum was posted online on October 15, 2012 on the Annual Reviews in Plant Biology web-site [3].

    Regarding the Annual Reviews erratum, how does one retract a figure and important conclusions in the text without disrupting the entire conclusion of the entire review? It is odd for the authors to claim that “The authors … wish to point out that other data and publications discussed in the Annu. Rev. Plant Biol. … article are not affected by the retraction of the PLoS ONE publication…” Is perhaps Annual Reviews trying to protect the image of this journal which, in plant science, ranks No. 1, with a 2012 Impact Factor of 23.654 by simply issuing an erratum, and not a retraction?

    Next, I paste the erratum notice of the Annual Reviews paper in its entirety:
    “Figure 3 and one paragraph on page 259 (text below) described the production and accumulation of protein bodies in transgenic cassava plants expressing a chimeric protein named zeolin. Figure 3 and the paragraph refer to a publication in PLoS ONE (Abhary M, Siritunga D, Stevens G, Taylor NJ, Fauquet CM (2011) Transgenic Biofortification of the Starchy Staple Cassava (Manihot esculenta) Generates a Novel Sink for Protein. PLoS ONE 6(1): e16256). The authors of the PLoS ONE publication have been unable to confirm the presence of the zeolin gene within the transgenic cassava plants. This raised concerns about the validity of the results reported in the article, which resulted in retraction ( Consequently Figure 3 and the paragraph on page 259 in Annu. Rev. Plant Biol. 2011. 62:251-272 are also retracted. Text of the paragraph to be retracted: An alternative strategy to enhance protein accumulation in cassava roots was to express chimeric storage proteins that were designed to form protein bodies, which would accumulate in the endoplasmic reticulum. The initial chimeric storage protein tested was zeolin expressed under control of the patatin promoter. Zeolin is a fusion protein of phaseolin from Phaseolus vulgaris and gamma zein from Zea mays. Previous studies showed that zeolin accumulates in the ER as protein bodies (e.g., 77). Total protein content of storage roots harvested from zeolin-expressing plants reached between 10% and 12.5% dry weight, an approximately fourfold increase (Abhary et al. 2011). This additional protein was stored as protein storage bodies (Figure 3). The authors apologize to the readers and wish to point out that other data and publications discussed in the Annu. Rev. Plant Biol. 2011. 62:251-272 article are not affected by the retraction of the PLoS ONE publication cited above. This erratum was posted online on October 15, 2012.”

    Xiangkai Li, Jian Yang, Mark Manary, Kendal D Hirschi (2012) Biotechnologically-Modified Cassava: Protein Absorption Relative to Casein. J Bioequiv Availab 2012, 4:4.

  3. Submitted on 2014/09/26 at 1:20 am | In reply to JATdS; reposted after site migration

    I wish to provide the second update related to the authors of the retracted PLOS ONE paper. The statements about published papers are based on searches on PubMed, Elsevier’s + Scopus, Springer’s SpringerLink, Google + Google Scholar. Taylor and Francis and Wiley data-bases were not checked yet. I have found the following, but the Donald Danforth Plant Science Center (DDPSC) is welcome to correct or add any additional information publically.

    1) No trace of Mohammad Abhary can be found, no papers published after the 2011 retraction. What is interesting is that he obtained his Doctor of Philosophy in Biology at the University of Missouri, St. Louis and that Claude Fauquet and Sam Wang were his co-advisors [1]. The abstract that appears online: “In this thesis, methodologies of improving the protein content in cassava storage roots are demonstrated. Tissue and subcellular-targeted protein expressions are found to concentrate the expressed proteins and improve the amino acid profile in cassava storage roots. Fusing targeting signal peptides, which stabilize the expressed proteins in protective subcellular environments, enhance the storage protein accumulation in cassava root cells. In addition, establishing a strong protein sink in the storage roots redirects the nitrogen flow from cyanogens biosynthesis towards the protein synthesis machinery. On the other hand, the ability to increase the protein content indirectly by modifying metabolic pathways illustrates the huge potentials of cassava and the richness of its genetic resources. All this has been studied in field trials, where transgenic plants are evaluated for their phenotype and yield productivity. The objective of all the experiments performed in this study is to increase the protein content in cassava storage roots and provide solutions to the malnutrition problems for those who cannot afford to supplement their cassava meals with other nutritious eats.” There is no e-mail for this individual to be found anywhere. This is unfortunate because it would be important to get his commentary on the case. I thus call on the University of Missouri, St. Louis to publically release a full transcript of that thesis so that the link between his research conducted there, under the supervision of at least two DDPSC specialists (Fauquet and Wang), can be carefully analyzed, including its links to work conducted at DDPSC. At the moment, the border that differentiates the work that was conducted at the University of Missouri, St. Louis and that that was conducted at DDPSC is extremely fuzzy or possibly even non-existent. The link between Abhary and the DDPSC thus appears to span quite a few years, even during his stay at the University of Missouri, St. Louis simply because his two main supervisors were from DDPSC. Dr. Abhary obtained his MSc and BSc degrees at Al-Balqa` Applied University and Al-AlBayt University, respectively, both in Jordan.

    2) Dimuth Siritunga continues to be extremely active in publishing, and remains at the Department of Biology, University of Puerto Rico-Mayaguez, Mayaguez, Puerto Rico [2]. He has at least 7 papers in Springer, none in Elsevier journals, and he shares authorship with Richard Sayre, Claude Fauquet and others in the Annual Reviews of Plant Biology paper that only suffered minor damage (i.e., an erratum). Dr. Siritunga on his web site continues to list the retracted paper as if it still was published (click on the button “Ver todas” = See all). He has an e-mail, so he will be contacted once the RW story has been fully updated.

    3) Gene Stevens, also known formally as William (Gene) Stevens on his web-site, continues at the University of Missouri-Delta Center, Portageville, Missouri, USA [3]. Dr. Stevens fails to show a full list of his publications, including the retracted paper. He has an e-mail, so he will be contacted once the RW story has been fully updated.

    4) Nigel J. Taylor, who appears to be a very major and central player in most of the papers that emerged from the International Laboratory for Tropical Agricultural Biotechnology (ILTAB), DDPSC, remains on at DDPSC [4]. More on the VIRCA and BC Plus projects in a future mini-analysis, but it is important to note that Dr. Taylor heads the VIRCA project and is deeply involved in the BC Plus project, too [5]. Moreover, Dr. Taylor is actively publishing and continues to be a central figure at DDPSC, even though he was a key member of the now-retracted paper. Why has Dr. Taylor, a highly experienced tissue culture and transformation scientist, not publically addressed what went wrong in his laboratory? IT is unclear why Abhary seems to be taking the blame while the directly responsible supervisors, Fauquet and Taylor, have not provided any public explanation. In fact, it is unclear why the then (and still now) President of DDPSC, Dr. James Carrington, were kind enough to offer an explanation in the name of DDPSC, but why Fauquet and Taylor, the directly responsible for the transformation laboratory, could / did not step forward publically with an explanation. Taylor also fails to list a complete list of his publications.

    5) Claude M. Fauquet is most likely the most central personality of this case, but from whom we have heard nothing from. At the time that paper was published (and retracted), Dr. Fauquet was the Director of ILTAB and the Principal Investigator of DDPSC since 1999 [6]. Now, as executive director of Global Cassava Partnership for the 21st Century, Dr. Fauquet travels the world and is actively promoting his genetic cassava GCP21 program [7] as a solution to global famine, or at least in Africa, and he has published at least one high-level paper in a Springer journal, Food Security [8]. That paper has 30 authors. One is Larry Beach (author 4), who at one point seems to have been working at DDPSC [9]. The paper lists Dr. Fauquet at CIAT, Cali, Colombia. Thus the timing of Dr. Fauquet leaving DDPSC and the exact timing of him joining CIAT is very important. A key question is why has Dr. Fauquet, most likely the most central payer of this saga, stayed silent and not provided an explanation? I believe that Dr. Carrington needs to explain clearly publically why Fauquet left but Taylor stayed on at DDPSC. Clearly the 2014 paper needs to be explored in more detail, but a link to that analysis will appear later since the analysis is complex. Incidentally, Fauquet was given a knighthood “Chevalier de l’Ordre des Palmes Académiques” by the French Minister of Higher Education and Research in 2007 [10].

    [8]; (open access)

  4. Submitted on 2014/09/26 at 2:44 am | In reply to JATdS; reposted after site migration

    “Regarding the Annual Reviews erratum, how does one retract a figure and important conclusions in the text without disrupting the entire conclusion of the entire review?”

    Simple – it wasn’t an important conclusion of the entire review, merely a paragraph pointing out an alternative strategy. In fact, there wasn’t even a conclusion of the review – it does mention future directions but the now-retracted alternative strategy isn’t specifically mentioned nor is it implicit in any of those future directions.

    Please let us not go overboard with retractions, please.

    That the OMICS paper isn’t retracted yet…well…plenty of potential explanations for that, ranging from the Chinese authors not willing to retract to the journal itself not willing to do anything. In both cases Hirschi can jump up and down and left and right, but there’s nothing he can do. I myself have encountered a situation where the corresponding author wanted to retract, but co-authors did not, and thus the paper still stands, even though there now is an “addition” that points out the whole paper (not mine) is actually based on an artifact. Obviously the journal didn’t want to retract either.

    A small P.S: Sayre is currently at LANL – New Mexico Consortium. There is thus nothing odd about him no longer being listed at Danforth: he doesn’t work there anymore!

  5. Submitted on 2014/09/26 at 8:45 am | In reply to Marco; reposted after site migration

    Marco, good to have you in the discussion and equally good questions. I hope that others will follow because there are a lot of questions about this case and its downstream effects that remain totally unknown, or unclear. Regarding the retraction of one figure, or one chunk of text, where does one draw the line? Is two figures or three, or four “enough” or “not enough” for a paper to be retracted? Is the “retraction” of two figures and two chunks of text considered to be borderline? Do you see what I’m saying? This is not about going overboard with retractions, it’s about setting logical limits to what is sensical, or not, about what can be retracted. At least the ARPB issued an erratum, so it’s a good start.

    Regarding the OMICS paper not having been retracted, even though Hirschi said that was the plan, we have witnessed plenty of retraction notices here at RW in which one or more authors did NOT agree with the retraction, but the paper was still retracted. If the publisher considers this to be a serious issue, in line with what Hirsch had discovered, then it would have already retracted the paper. Apparently, the use of faulty plant material is considered to be acceptable by Hirschi’s co-authors, or by Omics. Either way, this reflects something unhealthy on the under-surface of this case, and the fact that all proponents have just left sleeping dogs lie, without any explanations, is of concern. The only people who can clearly explain what happened are Hirschi and his co-authors. I will contact Hirschi and his co-authors myself and request a public statement once I have added the other updates to this case.

    Thanks for the update on Dr. Sayre. His opinion later on will also be valuable as I also request him to weigh in, so nice to know where he is. But, kindly note, I was not being critical of Sayre. He did the right thing and got an erratum issued, although he too has failed to indicate publically how things work in the Danforth Plant Center laboratories in terms of background checks on vectors and “transgenic” material. And yes, I simply wanted to state that Sayre is no longer at Danforth, but that he was, without absolutely any insinuations.

    So why is this case so important? Firstly, the Danforth Plant Center as well as Fauquet are actively promoting transgenic cassava as the solution to Africa’s famine-related problems. And the former is receiving funding from The Bill and Melinda Gates Foundation, Monsanto and others. So, this is not some small or innocent project or objective. It is a long-term project that has massive political, economic and scientific consequences, and thus deep and careful analysis of all angles (and individuals) is required. For example, it is totally unimaginable that the transgenic nature of vectors and plant materials in a transgenic cassava laboratory were not verified. And despite the kindness of Dr. Carrington coming forward with an immediate reaction, he and The Danforth Plant Center have still not provided any logical, transparent or viable explanation about how Taylor and Fauquet’s laboratory failed in confirming the transgenic nature of that specific “transgenic” cassava. That means, how much “transgenic” cassava carrying other genes that emerged from that laboratory was actually transgenic, and how much was not? How will we ever know? There are simple techniques that can be employed to verify transgenes (in vectors and plant material) such as PCR, RT-PCR, Southern, etc. that already have cheap kits on the market that allow for relatively easy verification. These techniques have been around for as long as ILTAB existed, so there is no “technical” or “economic” excuse. So, to put this retraction into a wider perspective, how can we trust that “transgenic” material claimed by the Danforth Plant Research Center in any of its papers (except where they show the transgenic nature using the techniques I describe?), including in reviews, is in fact transgenic? In other words, every time the literature talks about a “transgenic cassava plant” produced by ILTAB or by VIRCA and BC Plus projects is truly “transgenic”? Can you see the potential wide political, economical and social ramifications that this serious doubt is causing, and appreciate why this case may potentially be the most serious one seen in plant science simply because of the deep doubts of research integrity and/or competence at the Danforth Plant Center caused by this retraction and the potential serious ramifications?

  6. Submitted on 2014/09/26 at 9:51 am | In reply to JATdS; reposted following site migration

    I wish to focus on the VIRCA and BioCassava Plus projects now. There are some facts and some concerns about the VIRCA web-page, headed by Nigel Taylor, one of the co-authors of the retracted PLOS ONE paper, and who naturally supervised the project that led to that publication, and was thus also responsible for the vectors, plant germplasm and the entire functionality of the laboratory. This is in essence the same laboratory (but with a different name, changed from ILTAB to VIRCA) that is continuing to do essentially the same research, and publish papers subsequent to the retracted one.

    1) Incomplete funding statements. The web-site states: “The VIRCA project is funded by the Bill & Melinda Gates Foundation, the Unites States Agency for International Development from the American People (USAID), and the Monsanto Fund.” But the PDF file states [1]: “The VIRCA project is a collaborative effort between KARI, NaCRRI and the Danforth Center. The project is supported by Uganda and Kenya Governments, the Bill & Melinda Gates Foundation, the United States Agency for International Development, the Monsanto Fund and the Howard Buffett Foundation.”

    2) The link between VIRCA and the BioCassava Plus [2] project is unclear, although it is evident that the working team of the former provides the fundamental support for the latter, and an intricate link with it, even though the web-site claims that there are only two researchers for such a massive project [3]. DDPSC should clarify this relationship, although an analysis of the “participants” [4] indicates the massive overlap between both projects. What is curious to note is that “The Bill & Melinda Gates Foundation funds the vital work of the BC Plus project.” What is unclear from the VIRCA and BC Plus pages is whether the Bill & Melinda Gates Foundation provides separate funding for EACH project, or one lump sum amount for broadly “cassava improvement” at DDPSC. Here, too, the amount of project funding by all funders and the discrimination of funding between these two core projects requires more transparency and detail. This because of the massive economic ramifications this project and “transgenic” cassava has, not only in the US in terms of gathering funding, but also in terms of the trickle-down effect on the economic well-being of the very African’s this group wants to help. In that sense, how exactly has this amount of money, in dollar amounts (or in terms of transgenic cassava tonnage), returned to Africa, the so-called charitable target of their potentially multi-million dollar projects? And is this the “product” that Claude Maurice Fauquet alludes to in his video [5] in which he states “we use the viral genes… we put them into cassava and doing this we create what we call virus resistant cassava plant … We hope that by 2016 we will deliver the first product in East Africa… that is absolutely free”? Other insight about Fauquet’s rationale can be found in a couple of YouTube videos (see other links in [5]).

    3) Why does the publications page of VIRCA only list 5 published papers, all from 2012, and none from 2013 and 2014 [6]? Furthermore, why does the BioCassava Plus page not list any publications, and why can a search for ILTAB in the 21-document library lead to absolutely no information about the ILTAB-derived publications? Why has DDPSC provided ZERO public explanation about this PLOS ONE retraction on its web-site?

    4) Why does the Collaborating Institutions page [7] list only three main collaborating institutions (The Donald Danforth Plant Science Center, “a nonprofit research institute”; The Kenya Agricultural Research Institute (KARI); The National Root Crop Research Institute in Nigeria) while the Participants page [4] lists more (in addition to those listed, University of Nebraska-Lincoln, University of Puerto Rico, Washington University School of Medicine, Baylor College of Medicine, Boyce Thompson Institute for Plant Research)?

    5) What is the exact function of these two science advisors (Larry R. Beach; Eugene R. Terry) [8]? See my comment about Larry Beach above.

    6) Are any of these personnel in the two projects paid? What grants, salaries or other benefits (including travel) are they given directly by the Bill & Melinda Gates Foundation and the other grand sponsors? Any why is there zero information about this on the web-page? I ask these questions because it would be important to know the ratio of funding received to the amount of investment that has actually trickled back to Africa (in real US$ amounts), the true and intended recipient of these “transgenic” cassava plants.

    7) What is Monsanto’s stake in this whole project and in transgenic cassava?

    [5] ; ; ; (adding Claude M. Fauquet video as the key words to Google or Yahoo reveals a lot of the massive drive being made to impose cassava as the world’s solution to Africa’s famine).

  7. Submitted on 2014/09/26 at 1:53 pm | In reply to JATdS; reposted following site migration

    I can only comment on two things:
    1) it is not possible to give a firm guideline on how much should be wrong to warrant a retraction, and when a correction is enough. In this case there simply is no case for a retraction, however. We will have to live with case-by-case decisions, just like in every part of life.

    2) Fauquet (and Wang) obviously relied on his student – a wrong decision, but that’s life. To extrapolate this one case to express doubt about all their work is just plain malicious in my opinion, especially since they took direct action with this paper. *They* found the problem, *they* did the investigation, *they* reported it to all relevant people. No one would have found it was wrong, if they had not taken action themselves. No one.

    If anything, this shows shows we do not need to have any doubt about the research integrity at Danforth, nor do we need to doubt their competence.

  8. Submitted on 2014/09/26 at 2:39 pm | In reply to Marco; reposted following site migration

    I wish to focus briefly on the authorship roles and responsibilities, which may address Marco’s comments in 2) that the authors have acted responsibly and that this is a closed case. The now retracted PLOS ONE paper describes these functions as: “Conceived and designed the experiments: CMF MA NJT. Performed the experiments: MA DS GS. Analyzed the data: MA CMF DS NJT. Wrote the paper: MA NJT CMF. Generated the greenhouse and field experiments: DS GS. Produced the transgenic cassava plants: NJT.”

    In terms of the ICMJE’s definitions (old and new ones; new = post July 29, 2014 [1]), i believe that several of these authors would actually not qualify as authors, specifically DS, who was not involved with paper writing. Re: the last phrase: Nigel J. Taylor produced the transgenic cassava plants. Since we now know that those so-called transgenic plants were in fact not transgenic, should then not the bulk of responsibility lie on Dr. Taylor’s shoulders? As I say, this authorship and responsibility declaration, which must have been devised by MA, NJT and CMF (i.e., Abhary, Talor and Fauquet), who wrote the paper, now brings into serious question the actual responsibilities (and possibly even functions) of all these individuals. Although some readers may feel that I am slaying a dead horse, please think about it: if this team of professionals (because their CVs reveal that they are certainly not amateurs) was unable to verify that their “transgenic” plant material was in fact “transgenic”, then how sure are we that other “transgenic” plants produced by Taylor (and the Danforth Plant Center) before and after the PLOS ONE paper were in fact de facto “transgenic” plants?

  9. Submitted on 2014/09/26 at 3:58 pm | In reply to JATdS; reposted following site migration

    1) Writing =/ critically revising for important content. You will find loads and loads of papers in PLoS One in which some of the authors did not “write the paper”. I myself am on papers where I have made important modifications and improvements of the text, but I would not say I “wrote the paper”. According to the ICMJE guidelines I therefore would be considered an author

    2) Making the initial experiments to prepare the transgenic GMO and making sure the plants are transgenic have a lot of steps in-between. You can hardly blame Taylor if Abhary made up stuff that made Taylor believe he indeed produced transgenic plants

    3) Obviously Danforth does follow-up experiments, or they would not have detected these particular GMO plants were actually not zeolin-producing. I therefore have very little doubt those others *are* transgenic plants, as Danforth clearly does follow-up experiments

  10. Submitted on 2014/09/27 at 1:37 am | In reply to Marco; reposted following site migration

    1) Exactly! And since DS did not write, he also appears to not have checked the paper critically for important content. That makes him both (potentially) a non-author, and an irresponsible one at that, especially if we consider the new ICMJE definitions that state that ALL authors MUST take public responsibility for the paper, its content, its results and thus its downstream effects. It would be nice, but likely improbable, if COPE could offer comment about this particular case.

    2) Where is the proof that Abhary has “made this stuff up” outside of the investigation reported here? Abhary cannot be traced, his e-mail is non-existent and the only proof we have of what actually happened in that lab is from Dr. Carrington, and not from Dr. Taylor or Dr. Fauquet. To date, we have only heard ONE voice from Danforth Plant Science Center. Still, independent of who physically made the errors, as you correctly state, there are many steps in between. I know that from experience. Which is precisely why it is highly improbable that the errors lie exclusively in one individual’s court, because checks are constantly required for transgenic material. Finally, I want to emphasize that ultimately, no matter who made the error, that the directly responsible ones are only two individuals, the senior author and the PI of that laboratory / paper, i.e., Taylor and Fauquet. And, the ultimate responsibility is of course, Dr. Carrington, who represents the overall voice of the Danforth Plant Science Center. If you do not have a vertical chain of responsibility to match the vertical chain of power, then there is something chronically wrong with your laboratory / institutional structure.

    3) Where is the proof that the Danforth Plant Science Center has done follow-up experiments for ALL its transgenic material? What about all of the “transgenic” material that appeared in papers from the 1990’s until 2012? How do we TRULY know that that was really “transgenic” cassava? As I briefly indicated, given the very important funding received by the cassava group at the Danforth Plant Science Center, and the very possible political, economic and social implications of this whole cassava biotechnology programme, my concerns should not be taken lightly and every effort should be made by the Danforth Plant Science Center to provide irrefutable proof that every single “transgenic” plant, and not only the (non) zeolin-producing plants, were truly transgenic. I believe that you are wrong to put the ball into the court of the public. The onus of responsibility lies squarely on the shoulders of the Danforth Plant Science Center, and that is one objective of reviving this case: to get a public explanation for so many aspects that appear to me to be unclear, or incomplete.

  11. Submitted on 2014/09/27 at 2:22 pm | In reply to JATdS; reposted following site migration

    Now finally, we have reached the crux of the matter. I wish to now focus on one of the most important aspects of this case, and indicate why we should pursue as deep and thorough analysis of all parameters as possible, perhaps in a way in which no other retraction case has yet been analyzed at RW. Access to publically available data gives us a better picture about what the downstream effects of a retraction might be, at least in terms of publishing and publications. So, in a sense, this Danforth Plant Science Center-related case could be considered a type of model case for RW. The suggested initial argument / question / concern posed implicitly in the original title of the RW story was: does the PLOS ONE retraction have a domino effect, or any downstream effects? So how does one go about assessing a downstream effect? I have done my best to analyze the literature related to this publication. Even so, I feel that the analysis may be quite preliminary, but it is a good start to initiate deeper discussion, and clarification. So, greater openness and transparency from the Donald Danforth Plant Science Center as well as open communication by the authors of papers that have referenced the retracted PLOS ONE paper would be greatly appreciated. I am hopeful that Dr. Carrington (or other DDPSC members) will voluntarily come forward to address the several gaps and concerns I have about this case and its actual and possible downstream effects.

    First, the facts (obtained from the PLOS ONE web-page), true up to September 26, 2014:

    1) Total article views: 10360, including 1292 PDF downloads;

    2) If one observes the plotted graph of views, the retracted paper continues to be viewed, as indicated by the cumulative monthly usage.

    3) The PDF file, as well as the HTML text of the retracted paper remain intact, and other than a red box with the notice of “Retraction Notice”, the text, and most importantly the open access PDF file, remains without a red water-stamped PDF file. That means that PLOS ONE may be held directly responsible for the incorrect use of a retracted paper by the cassava research community or by plant researchers more widely. In other words, anyone who accesses the PDF file of this retracted paper, for example through a Google search, will NOT know that the paper is in fact retracted. They will not know that the results are false, that the “transgenic” plants claimed are in fact not “transgenic” and that the paper that they are reading, and possibly referencing, is false! The fact that PLOS ONE does not adopt a policy that, for example Elsevier adopts, namely to add the water-marked red stamp of RETRACTED across every page of the PDF file, is extremely worrisome. In this case, the publisher might actually be responsible for the incorrect proliferation of “retracted” references in the literature. This also indicates that the warning or safety net in place to warn potential readers that this paper is a red flag are insufficient at PLOS ONE (vs, for example, the system in place by Elsevier, which is far more robust and effective).

    4) To test the hypothesis proposed in 3), using ONLY this paper as the reference, one in fact can observe that the paper has been cited 11 times on Scopus (Elsevier), 10 times on CrossRef, 2 times on PubMed PMC, 8 times on ISI Web of Science, 6 times on Europe PubMed Central, and 29 times on Google. It is obvious that there are overlaps between these data bases, but the numbers that emerge after applying filters to check for overlap are in themselves astonishing, I believe. The numbers indicate that no less than 14 papers have referenced the retracted paper, although the exact number of papers that were published or that were in production before the actual retraction took place is not absolutely clear. One can however, most likely, safely assume that, given the actual date of the retraction (7 September, 2012), that about 13/14 of those papers that in fact did reference the now retracted paper did so before the retraction appeared. Nonetheless, there needs to be a retrospective correction of the literature, independent of whether the papers that referenced the retracted PLOS ONE paper were published before, during or after the actual retraction was issued. And this is precisely the reason why this case could serve as a brilliant model case for RW and retractions to understand the real effects of the downstream effects, precisely because we are able to quantify the publishing-related effects.

    The summary of those 14 papers that reference the retracted Abhary et al. PLOS ONE retracted paper is detailed next:

    1. The BioCassava Plus Program: Biofortification of Cassava for Sub-Saharan Africa Sayre R, Beeching J, Cahoon E, Egesi C, Fauquet C, Fellman J, Fregene M, Gruissem W, Mallowa S, Manary M, Maziya-Dixon B, Mbanaso A, Schachtman D, Siritunga D, Taylor N, Vanderschuren H, Zhang P. Annual Review of Plant Biology 2011 62(1): 251. doi:10.1146/annurev-arplant-042110-103751 (now with an erratum)
    2. Cassava Genetic Transformation and its Application in Breeding Liu J, Zheng Q, Ma Q, Gadidasu K, Zhang P. Journal of Integrative Plant Biology 2011 53(7): 552. doi:10.1111/j.1744-7909.2011.01048.x
    3. Overexpression of Hydroxynitrile Lyase in Cassava Roots Elevates Protein and Free Amino Acids while Reducing Residual Cyanogen Levels. Narayanan, N.N., Ihemere, U., Ellery, C., Sayre, R.T. 2011 PLoS ONE 6(7), e21996
    4. Efficient transformation and regeneration of transgenic cassava using the neomycin phosphotransferase gene as aminoglycoside resistance marker gene Niklaus M, Gruissem W, Vanderschuren HGM Crops 2011 2(3): 193. doi:10.4161/gmcr.2.3.18866
    5. Niklaus M, Gruissem W, Vanderschuren H. Efficient transformation and regeneration of transgenic cassava using the neomycin phosphotransferase gene as aminoglycoside resistance marker gene. GM Crops 2011; 2:193 – 200; PMID: 22179195; Of note, GM Crops and Food, a journal published by Landes Bioscience was recently transferred to Taylor and Francis, so I am hopeful that Taylor and Francis will take note of the error and issue an erratum.
    6. Robust transformation procedure for the production of transgenic farmer-preferred cassava landraces Zainuddin I, Schlegel K, Gruissem W, Vanderschuren H. Plant Methods 2012 8(1): 24. doi:10.1186/1746-4811-8-24
    7. Using storage organelles for the accumulation and encapsulation of recombinant proteins Khan I, Twyman R, Arcalis E, Stoger E. Biotechnology Journal 2012 7(9): 1099. doi:10.1002/biot.201100089
    8. Cassava about-FACE: Greater than expected yield stimulation of cassava (Manihot esculenta) by future CO2 levels Rosenthal D, Slattery R, Miller R, Grennan A, Cavagnaro T, Fauquet C, Gleadow R, Ort D. Global Change Biology 2012 18(8): 2661. doi:10.1111/j.1365-2486.2012.02726.x
    9. The formation, function and fate of protein storage compartments in seeds Ibl V, Stoger E. Protoplasma 2012 249(2): 379. doi:10.1007/s00709-011-0288-z
    10. Cloning and characterization of a tuberous root-specific promoter from cassava (Manihot esculenta Crantz) Koehorst-van Putten H, Wolters A, Pereira-Bertram I, Berg H, Krol A, Visser R. Planta 2012 236(6): 1955. doi:10.1007/s00425-012-1796-6
    11. Iron and protein biofortification of cassava: lessons learned Leyva-Guerrero E, Narayanan N, Ihemere U, Sayre R. Current Opinion in Biotechnology 2012 23(2): 257. doi:10.1016/j.copbio.2011.12.009
    12. Variations in the chemical composition of cassava (Manihot esculenta Crantz) leaves and roots as affected by genotypic and environmental variation Burns, A.E., Gleadow, R.M., Zacarias, A.M., Cuambe CE, Miller, R.E., Cavagnaro, T.R. 2012 Journal of Agricultural and Food Chemistry 60(19):4946-56. doi: 10.1021/jf2047288
    13. Biotechnology for Enhanced Nutritional Quality in Plants Uncu A, Doganlar S, Frary A. Critical Reviews in Plant Sciences 2013 32(5): 321. doi:10.1080/07352689.2013.781453

    Of course, we should not forget the 2012 Hirschi paper (the 14th) I describe above that not only references the retracted paper, but also used “transgenic” cassava material obtained from Fauquet. Finally, one more paper that strangely did not appear captured by the PLOS ONE hyperlinking functions was in

    Closer scrutiny of these 14 downstream papers that cited / referenced the PLOS ONE paper reveals the following:

    a) Why did Dr. Sayre, who apparently took responsible actions towards the ARPB paper (ref 1 above), not take ensuing responsibility towards references 3 and 11, of which he is an author? It is not as if he did not (and does not) know about the retraction, so why has he decided to leave the reference of a retracted paper from his own previous research institute (Danforth Plant Science Center)? Fauquet and Sayre’s departures happened during the post-retraction period. This then additional questions that the President, Dr. Carrington, now needs to address publically, simply because the official institute listed on those papers is precisely the Donald Danforth Plant Science Center. That means if Sayre and co-authors are unable to assume their academic responsibilities of requesting the publisher to issue an erratum, then the Danforth Plant Science Center has no option but to assume that responsibility, not only for the Sayre papers, but in fact for all 14 papers that reference the retracted PLOS ONE paper. As one example, and an equivalent measure of responsibility that the Danforth Plant Science Center should follow is by the Catholic University School of Medicine, which took the responsibility of requesting the journal to retract the erroneous papers of a now deceased researcher, Paola Palozza [1]. Similarly, Dr. Carrington surely has the responsibility of contacting the editors and the journals listed above to request, at minimum, an erratum, with a sincere apology to the readership and the scientific community.

    b) The same question should be posed to Dr. Fauquet regarding reference 8. Why would a former DDPSC PI that is now aware that his own paper has been retracted not contacted Global Change Biology within these two years to issue an erratum?

    c) It is extremely surprising to see some pretty big names in plant science like Zhang, Gruissem, Vanderschuren, Cavagnaro, Visser, and Frary reference a retracted paper. They too need to be contacted to understand if: i) they were not aware that the paper was retracted for the reasons I explain above; ii) they knew that the paper was retracted but simply decided to reference it anyway; iii) they plan to contact the journal / publisher to issue an erratum but have not yet done so. It is also extremely important for the authors of these papers, and as independent post-publication verification agents, the wider plant science community, to also assess whether reference to the retracted Abhary et al. PLOS ONE paper might in fact invalidate any claims that they have made. If we could get their responses, we may be able to start unlocking one of the black-boxes surrounding retractions, i.e., the reasons why any scientist would (want to) reference a retracted paper, except to focus on the retraction itself.

    d) What about secondary downstream effects. In other words, what about papers that reference papers that have incorrectly referenced a retracted paper? Should they also be the subject of errata or expression of concern? Allow me to exemplify. Please take reference 3), the Narayanan et al. paper (all those authors are incidentally from the Danforth Plant Science Center). Please click on the “Metrics” page of the PLOS ONE site. There you will see that 5 papers (Scopus) and 6 papers (ISI) have referenced this first domino paper, 2 of which already appear in the list of 14, but 3 of which are new, i.e., secondary dominoes. Should the authors of those secondary downstream papers / dominoes also be contacted to alert them of this problem? I believe yes, because at some point, the domino effect must stop. If authors of “secondary domino literature” are not alerted, and if those editors and journals are not alerted, then there is a serious possibility of deeply corrupting the literature indefinitely.

    e) Who should be responsible for alerting the journals / publishers that published these references to the retracted PLOS ONE paper? Why did the editors and EICs of those journals not detect the retracted paper (pre- or post-retraction) and say something during the peer review process (in the case of the former)? Could this be another sign of the failure of traditional peer review (in the former case)? I am of the opinion that three entities have the full responsibility here of issuing 14 errata (at least): i) Dr. Carrrington, as the maximum leader since the errors emanated from a laboratory in his research center under his supervision and command; ii) aiding Dr. Carrington should be Dr. Fauquet and Dr. Taylor, the direct PIs and senior researchers of the laboratory from which these false transgenic plants emerged; iii) the authors of the actual papers themselves, once they have been informed about this case and this update that I have provided at RW.

    So, what is, or should be, the action plan? In essence, what I am going to do is to contact Dr. Carrington and his staff at the Danforth Plant Science Center, the authors of all 14 downstream papers, as well as general cassava biotechnology specialists, and possibly even the editors-in-chief and editors of the relevant journals. This shows in fact how one retracted paper has now actually come to affect more than a dozen papers, directly affect several dozen researchers, many journals and publishers, data-bases, and even research institutes. In other words, the domino effect is not minor, it is quite MASSIVE when all these links are taken into consideration.

    The objectives of contacting these entities are ultimately to: a) obtain some clear responses from Dr. Carrington and all DDPSC-associated researchers (current or former) about gaps in the information and concerns that were not addressed for two years now; b) to get someone to take public responsibility; c) to get the wider cassava (and hopefully plant science) community more actively engaged in these issues that surround (or underlie) retractions. Without frank, open, honest and wide discussion, how are we ever supposed to know what went wrong, and more importantly, what is wrong with the (plant) science literature, and how we can correct it? Finally, at the end of the day, for this particular case, what we want to know is how “transgenically true” is all of the material that was claimed to be “transgenic” and published by any cassava researcher, particularly the “transgenic” plants that were produced by the Donald Danforth Plant Science Center?

  12. Submitted on 2014/09/27 at 4:21 pm | In reply to JATdS; reposted following site migration

    1) Good grief, what is it with your ICMJE *guidelines* as some kind of absolute rules? And who says he didn’t check the paper?
    2) Ask yourself why Danforth took action. I don’t think anyone would have ever noticed. They could even easily have said, later, that they could not repeat the experiments. Take that together, and I have absolutely no doubt there is one person who cooked the books. I know a lot of people here love to blame the PIs if one of their minions commit fraud, and I can see how you could put some level of ‘responsibility’ on them, but this is just too easy. There simply are bad people out there, at all levels, and I know how deceptive such people may be. Your ‘fault’ in that case is to trust people. As I said earlier on RW, I personally do not want to work in an environment where I constantly am asked to distrust the people I work with. And again, Danforth obviously took their responsibility, doing all the work to check what had happened, and retracting the paper.
    3) JATdS, tell me what “proof” would satisfy you. Also, tell me why Danforth should give in to *your* demands.

  13. Submitted on 2014/09/30 at 11:24 am | In reply to Marco; reposted following site migration

    1) You seem to have misread my interpretation of the ICMJE. My record will show how I have been consistently critical of the ICMJE and its imposition of “ethical” values on scientists. I am not against the logic of the ICMJE guidelines, because some of them actually make sense, even though in a real, practical situation, there are plenty of papers, and laboratories, in which one or small groups of individuals do separate but essential functions, but who are authors nonetheless even though they have not devised the intellectual seed of the project. However, I am strongly against the way in which guidelines are being imposed on scientists by publishers as if they are some sort of a law. But the ICMJE is not the focus here.

    2) I am of the opinion that the many unanswered questions remain unanswered with respect to this case. Once again, Marco, you claim that only one person “cooked the books”, yet you (and the Danforth Plant Science Center) have not yet to provided concrete evidence of this (only Dr. Carrington’s word). If you (plural) are so sure, then simply make the evidence public. Show us the lab records, the lab books, get Abhary to respond here, too, to get a balanced perspective, even if he is retired from science. Why has this story stayed silent for 2 years, with so many unanswered questions? Regarding DDPSC’s responsibilities, I believe that they have only dealt with one. They now have, as I already explained in considerable detail above, a number of remaining responsibilities to take care of and questions to respond to.

    3) I think the questions I asked above are fairly self-explanatory. What I will say is the following: why has no-one else in the plant sciences been asking these critical questions, and why has no-one seen or analyzed the literature and this case in any detail? It’s not about answering to me, it’s about answering to science, plant scientists and the public. The public funds the Gates Foundation, thus it is public / private money that funds that research institute, and so when basic things about a transgenic laboratory go wrong (like not having transgenic plants as they claim in a paper), then yes, the reputation of that institute must be questioned, as must the research that has emerged from it. If they had received no money, then I agree that critical analysis could perhaps be more tempered, but it is precisely because of the high-profile nature of this cassava project that it needs an absolutely critical analysis. Kindly note that I have added only questions over the past two days, and so we need to give time for responses to filter back. Those responses will hopefully say something, as equally as will silence.

  14. Long awaited domino (perhaps Lego) effect of retracted papers. Thinking of meta analyses.
    No words can describe how I feel.

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.