For months, a researcher has wrestled with a journal over the wording of an upcoming retraction notice. It appears that she has lost.
Earlier this week, Cell retracted the paper, despite the protests of first author Shalon Babbitt Ledbetter. When Ledbetter learned the journal was planning to retract the biochemistry paper over image manipulations, but wouldn’t name the culprit in the notice, she shared her concerns on PubPeer. Although a 2015 letter sent to Cell from Saint Louis University identified last author Dorota Skowyra as responsible for multiple manipulations, the journal wasn’t planning to say Skowyra was responsible in the retraction notice. Which would leave all other authors — particularly Ledbetter — under a cloud of suspicion.
Now, Cell Press has finally retracted the paper, along with another paper in Molecular Cell that lists Skowyra as corresponding author. Both notices describe image manipulations that were investigated by Saint Louis University (SLU). Neither identifies who is responsible.
Ledbetter and the first author of the Molecular Cell paper both told Retraction Watch they were disappointed with the journal’s decision to retract their papers. Ledbetter also reiterated her concern that the journal’s decision not to identify who was responsible for the manipulations could hurt her, as first author:
I viewed this selective omission on their part as both unfair and deceptive, especially since Cell had official documentation from SLU providing accountability for the alleged data manipulations uncovered in the paper. But Cell chose to proceed with the version of the notice they crafted, and there was nothing I could do to convince them otherwise.
A spokesperson for Cell Press declined to comment. At press time, Skowyra told us she was working on a response to our questions with her coauthors. The two retracted papers share three authors in common: Skowyra as last and corresponding author, along with two middle authors.
Last month, a spokesperson for SLU sent us a statement from Matthew Christian, associate vice president for research, saying that “a SLU author” of the Cell paper was investigated and “was cleared of research misconduct by SLU and [Department of Health and Human Services].” That statement appeared to contradict the 2015 letter to Cell from SLU, which noted:
…the committee found that Dr. Skowyra did engage in image manipulation as alleged, a practice that they deemed consistent with research misconduct…
When we contacted SLU about the two retractions, we received another statement from Christian:
Saint Louis University has been notified by Molecular Cell about its April 19, 2018 retraction of a 2006 article co-authored by a SLU scientist. In 2014 , SLU’s Research Integrity Office thoroughly investigated concerns raised about the article by a journal reader and promptly reported its findings to Molecular Cell. SLU’s investigation found cosmetic changes were made to two images in the paper but did not find manipulation of the underlying data. Further, SLU’s findings cleared the scientist of research misconduct, as did an independent investigation by the National Institutes of Health’s Office of Research Integrity. We respect the journal’s right to make an editorial retraction on a paper, according to its standards.
We know much of the backstory behind the Cell retraction — in March, we spoke to Ledbetter, who forwarded us email correspondence with the journal that included draft text of the upcoming notice. The published text is somewhat different — there are some wording changes, and the final text specifies that “the manipulations did not affect conclusions drawn from the data.”
Ledbetter told us:
In regards to my agreement or disagreement of the retraction itself, I have always objected to the retraction of this paper on the grounds that the science was and still is sound. Although I could appreciate the position of the Cell editors in wanting to distance themselves from any appearance of being complicit in any alleged impropriety by allowing the paper to remain as originally published, a retraction of the article from the scientific record gives the appearance that the data and conclusions in the paper were fundamentally invalid, which is simply not true.
Cell continually made the claim that they needed to “correct the scientific record,” and to do so, retraction of the paper was the only reasonable course of action. However, despite these claims, Cell continually delayed the retraction. I first learned from Cell in November 2017 that they were retracting this paper. In the few months I had to gather the facts, I later learned that this ordeal with Cell had been going on since 2014. This extensive delay by Cell to retract this paper leaves me unconvinced about their purported motives to “correct the scientific record.”
Here’s the retraction notice for “Destabilization of Binding to Cofactors and SCFMet30 Is the Rate-Limiting Regulatory Step in Degradation of Polyubiquitinated Met4,” published by Molecular Cell:
Molecular Cell is retracting this article. In response to concerns raised by a reader, we contacted the corresponding author for original data and investigated the issues. The Research Integrity Office at Saint Louis University also investigated the matter. Both investigations determined that data in two figures had been inappropriately manipulated. In Figure 4A, a section from one part of the image was duplicated and placed over another part of the image. In this instance the authors could not locate the original data; thus we cannot determine whether the alteration obscured any relevant, underlying data. In Figure 5B there is a splice between lanes 2 and 3. While the splice in itself does not necessarily indicate inappropriate manipulation, the investigations found that what is depicted in lane 3 is not what appears in the original data image. Additionally, in Figure 5B, one of the lanes (labeled “alphas”) was inappropriately manipulated to remove blemishes. Given these issues, we are retracting the paper. The authors do not agree to the Retraction.
The 2006 paper has been cited 23 times, according to Clarivate Analytics’ Web of Science.
The first author of the Molecular Cell paper, Srikripa Chandrasekaran, told us:
I want to state that I was not involved in alterations of the images in question. I wish those image changes could have been spotted either by me, other coauthors, reviewers, or editors of cell press before the manuscript was accepted for publication. In my opinion, while the image adjustments should not have taken place, they have no bearing on the scientific validity of the paper, and it would have been a better course of action from the journal’s perspective to allow an erratum of this work.
In 2016, the journal Cell Division published an erratum for a paper by Skowyra, also citing “unacknowledged modifications” to images.
Like Retraction Watch? You can make a tax-deductible contribution to support our growth, follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up for an email every time there’s a new post (look for the “follow” button at the lower right part of your screen), or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
No matter how unfair it may seem, I was always of the opinion that the buck stops with the primary/ corresponding author. If there was impropriety of one of the other authors and the paper has to be retracted, then the corresponding author is at least partially responsible. Similar to the responsibility of the PI of a funding source for the integrity of the data produced. I have been a PI, primary and corresponding author, and a co-author and have made certain that everything published under my name in any context is rigorous. However, I do understand that that is easier said than done. What do other authors on here think about your responsibilities as corresponding author, and how you do or will ensure the integrity of the data you publish?
It would be nice if senior authors, mentors, corresponding authors, etc., could do what you suggest. And I believe they should bear some of the responsibility for funny business that occurs during their watch. However, only those who actually did the misconduct are (and should be) responsible for it. So, maybe a PI is responsible for letting misconduct through, but he/she is NOT responsible for DOING the misconduct (unless he/she actually did it or was complicit).
Further, the way things are now, often there’s no way for every coauthor to view each datum, gel, calculation, etc. And even if one was able, the interdisciplinary nature of science is such that it’s highly collaborative and specialized. It’s often true that every author is not expert in each experimental approach/method, creating a vulnerability for coauthors of many works. Blame should remain where it belongs–with the cheaters. Maybe there should also be some level of punishment for inattentive and unwitting accomplices, but that’s a separate issue.
Finally, as I believe RW and its readers have said in the past: the process should be more open and transparent and speedy. And when a conclusion is reached, it should be disseminated as widely as possible.
But it’s the corresponding author, who is (allegedly) responsible for the image manipulations in this case. As far as I recall, the first author was a research technician at the time she carried out the work leading to the Cell paper. Based on a quick pubmed search, this was her very first publication. Hard to blame her for assuming that her PI will handle her data appropriately.
If the corresponding author engaged in misconduct, then that’s the person who should “face the music”. Again: whoever engages in misconduct should be the person(s) against whom findings should be made.
All authors gain from the publication….
Fraud and misconduct by specific individuals must be dealt with individually. Retraction is not that mechanism.
It is not clear to me that a retraction was the appropriate outcome from this description. If the image manipulations did not affect the conclusions, then the paper should not be retracted. The experiments should be repeated and the figures corrected with data that were not manipulated. The intent of a retraction is to remove from the published record a paper with sound conclusions and seems punitive rather than attempting to advance scientific progress and correct the published record.
When a biology paper is retracted or corrected, the author often mentions that the problematic image does not affect the conclusions. This is ridiculous. If it does not affect the conclusions, why was the image included in the paper in the first place? IN mathematics, this is considered to give incorrect proof to something true, which taking over the chance of others who works on the correct poof.
So, it has to be retracted.
I don’t think that’s a completely accurate account of the practice in mathematics. Many incorrect proofs of true statements have been published (it has been said of Solomon Lefschetz that he never published a false theorem, and never gave a correct proof…). As long as the “true statement” remains unused by the rest of the mathematical community (a fate which I think is still common, though not as common as it used to be not so many decades ago, when the median number of papers published by an American Ph.D. in mathematics during his or her entire career was one; that suggests strongly that, at least, any previously unstated “true statements” in those singleton papers were never used later by anyone, including their authors), its publication can cause no serious harm. Conversely, other mathematicians who find the “true statement” and want to make use of it will usually (in my experience) try to verify it, usually (at first) by trying to understand the purported published proof: if errors are detected in that proof, but in the process of finding them a new, correct proof is developed, there is usually no obstacle (always assuming the “true statement” is actually of interest to anyone) to publishing the new proof (and incidentally pointing out the errors in the old one).
I have been personally involved in one such case. In 1957, the eminent Italian algebraic geometer Beniamino Segre published a purported proof that, if polynomial functions p(z) and q(z) of a single complex variable “separate points” of the complex numbers (i.e., if there do not exist two distinct complex numbers z0 and z1 such that p(z0)=p(z1) and q(z0)=q(z1)), then either the degree of p divides the degree of q or vice versa. In 1970, the Mexican algebraist Ignacio Canals and algebraic topologist Emilio Lluis published a paper pointing out the error in Segre’s proof, and giving a different proof. In 1974 the Japanese complex analyst Masakazu Suzuki pointed out the error in Canals and Lluis’s proof, and gave a (correct!) analytic proof; more or less simultaneously, the algebraic geometers Shreeram Abhyankar and Tzuong-Tsieng Moh (respectively Indian and Chinese) gave a (correct!) algebraic proof (using only what Abhyankar was pleased to call “high school algebra”, i.e., polynomials, as contrasted with “college algebra”, i.e., rings and fields, and “university algebra”, i.e., categories). In 1982 I published yet another proof, this one by (not quite “high-school”, but definitely not “university”) knot theory. Since then about 20 other pairwise distinct proofs have been published (I, of course, like mine the best…). Nobody, as far as I know, was harmed or held back (in the years between 1950 and 1974) by the publications of Segre or Canals and Lluis.
In mathematics, one can be famous by posting a conjecture, posting a hypothesis, or proving a conjecture.
But mathematics do not agree that a fact is true without a valid proof. A mathematician may or may not retract his incorrect proof. But it is funny or ridiculous if he claims that this theorem is true even his proof is wrong.