PLoS ONE has just issued a 12-figure correction on a paper by Mario A. Saad, who sued the American Diabetes Association unsuccessfully in an attempt to prevent it from retracting four papers in its flagship journal Diabetes.
The corrections include taking out Western blots copied from another Saad paper, as well as several figures where the bands were “misplaced.”
PubPeer commenters suggest this may not be enough, calling seven additional figures into question.
We’ve reached out to PLoS to find out why they didn’t just retract the paper. We also asked if there will be a correction or retraction for Saad’s PLoS Biology paper “Gut Microbiota Is a Key Modulator of Insulin Resistance in TLR 2 Knockout Mice,” which PubPeer users have tapped for multiple alleged image problems [see update at the bottom of the post].
Here’s the complete correction for “Atorvastatin Improves Survival in Septic Rats: Effect on Tissue Inflammatory Pathway and on Insulin Signaling,” including links to the raw images:
Several figures in the article included the wrong blots. The authors apologize for these errors and are providing corrected figures as well as the underlying raw blots.
The p-JNK blots in Fig. 4B were inadvertently include from lanes in Figure 3D of the publication below:
Diabetes. 2011 Mar;60(3):784–96. doi: 10.2337/db09-1907
Physical exercise reduces circulating lipopolysaccharide and TLR4 activation and improves insulin signaling in tissues of DIO rats.
Oliveira AG, Carvalho BM, Tobar N, Ropelle ER, Pauli JR, Bagarolli RA, Guadagnini D, Carvalheir JB, Saad MJ.
In Fig. 2B, 2C, 2G, 4D, 6G and 6I several images of the lower bands are misplaced. This affects the following panels:
Fig. 2B: beta actin blot
Fig. 2C: insulin receptor blot
Fig. 2G: beta actin blot
Fig. 4D: beta actin blot
Fig. 6G: beta actin blot
Fig. 6I: beta actin blot
In Fig. 4B, 4E, 4F, 4G, 4H, 5A and 5B several images of the upper bands are misplaced. This affects the following panels:
Fig. 4E: pcjun blot
Fig. 4F: pcjun blot
Fig. 4G: p-IRS1 serine 307 blot
Fig. 4H: p-IRS1 serine 307 blot
Fig. 5A: NFkappaB blot by the following one:
Fig. 5B: NFkappaB blot
The paper has been cited 11 times, according to Thomson Scientific’s Web of Knowledge.
We emailed Saad — who now has four expressions of concern, one retraction, and this correction — to ask what happened, and will update if we hear back.
Update 3:15 p.m. EST 3/5/15: PLoS ONE deputy editor Iratxe Puebla emailed us the following statement:
The decision to issue a Correction was reached on the basis of the evaluation on the extent to which the issues with the figures affected the findings of the study. Issuing a Correction at this point does not preclude further evaluation or issuing a Retraction at a later stage if additional evidence comes to light that calls into question the integrity of the work. PLOS ONE is not involved in a lawsuit involving Dr Saad.
She also pointed out that PLoS Biology is editorially independent, so PLoS ONE can’t comment. We’ve contacted that journal separately.
Update 10:45 a.m. EST 3/6/15: PLoS Biology chief editor Emma Ganley told us via email that they are “looking into” the Saad paper, but wouldn’t give us any other details:
The paper you’ve asked about is one that the PLOS Biology team is currently looking into. We can’t share any details and, as is standard, we are not willing to discuss this externally. As you are likely aware, journals have to follow a clear process when investigating issues on papers and at all the PLOS journals we follow the COPE guidelines. Out of respect for both the scientific process, and for all involved, these things take time.
Update: 7:52 p.m. ET 3/11/15: In response to a comment, we asked PLoS ONE deputy editor Puebla if Saad had ever threatened the journal with a lawsuit. “There was no discussion about a legal case in our correspondence with the authors,” Puebla responded.
Like Retraction Watch? Consider supporting our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, and sign up on our homepage for an email every time there’s a new post.
Is a 12 figure correction a record for retraction watch?
Yet again, this leaves the PLoS Biology paper hanging out there like a sore thumb, with the credibility of the whole PLoS enterprise seeming to decay by the hour. A serious question that must be asked at this stage, is whether PLoS chose the correction route in response to legal threats from Saad? As a >$20m/yr operation, they’re an attractive target for a lawsuit. Surely it can’t harm for a journalist to ask PLoS directly?
PLoS ONE certainly has a lot of ‘splainin’ to do, but what of Saad? We already knew he was litigious, but does he truly have no sense of decency? This is so far, far beyond shameful.
PLoS One didn’t retracted the paper because University investigation (http://retractionwatch.com/wp-content/uploads/2015/02/saad-unicamp-investigation-report.pdf) concluded that the results are reliable. This was achieved by “talking with the member of the group” and because “[the] article was extensively mentioned in the literature” (although the only group that did the same experiment reported contradictory findings). This is the conclusion signed by investigators Licio Velloso and Jose Gontijo.
Saad has 119 collaborations with Velloso and 17 with Gontijo
https://www.ncbi.nlm.nih.gov/pubmed/?term=velloso+saad
https://www.ncbi.nlm.nih.gov/pubmed/?term=gontijo+saad
Yes, the argument that the results are reliable because the paper was highly cited is an utterly appalling one. Citation is not verification – as thousands of citations for people like Diederik Stapel prove.
An average 6-year old could see through the logic flaws in this Correction.
Plos One should have issued an Expression of Concern exactly as Diabetes did. When faced with an extraordinary request to amend Figures in this way five years after publication, a journal surely must consider it their duty to work out if other papers from the same authors are under suspicion. Then they would discover the problem and reasonably determine that an EoC was indicated given that ‘an investigation is underway but a judgement will not be available for a considerable time’ (the university already having botched one investigation).
In the correction they say: ‘The authors apologize for these errors and are providing corrected figures as well as the underlying raw blots.’
But they don’t provide the underlying raw blots. They show six rather fuzzy blots without labels or legends with nothing to say what any of them mean. There should be 6-8 blots for every single experimental condition since there were 6-8 rats – THAT’S WHERE THE ERROR BARS COME FROM. To get all those bar charts and error bars there should be hundreds of blots. Where are they?
As I’ve written before, the journal and university should be asking whether the reason they fabricate their images from blots in other figures and other papers is because those rats never existed.
That’s why the correction is so farcical.
No way this correct the entire record. There are still many apparently duplicated (should I say misplaced?) bands, even in the corrected panels. For instance, at 4B (JNK total) of the corrected figure, there are 4 apparently duplicated, identical bands. How can I contact PLoS One regarding this?
You can leave a comment on the paper, but do note they will remove any comments that suggest impropriety.
The precedent set by PLoS is terrible and it sends the absolute wrong message to the community. That means that any of us can make at least 12 errors to any figure in any paper, past or future, and we can demand that they never retract our paper, based on the Saad paper. Thank you, PLoS, for lowering the bar and the standards of science publishing. I hope the authors had to pay an APC for the correction. This is not a scientific paper. It is a fragmented scientific patchwork. Ideally, what needs to happen is the original paper needs to be retracted, the new figures need to be inserted, the original retraction needs to be referenced, with an explanation, and the new paper needs to be subjected to a fresh peer review. The scientific literature is starting to look like a mine-field.
I agree Fig. 4B still looks problematic. But I wouldn’t bother with PLoS One – the lights are on but no one is in. This isn’t the first time they have allowed corrections of essentially this form:
http://journals.plos.org/plosone/article?id=10.1371/annotation/d4645c95-739f-4175-b753-b6fe0ba6063c
The only action that works is to list the flaws in a high profile place that wakes the journals up to the problem. PubPeer doesn’t seem to be carrying enough weight. The best I’ve seen was 11jigen who produced amazing You Tube videos showing the fraud.
But who has time to do that kind of police work, make the video and upload it to YT?
The effect of pubpeer is rapidly wearing off. Responses from authors have become extremely rare. The typical scenario is that someone points out “irregularities”, a handful of peers occassionally express their dismay, then the comment eventually descends into the abyss of ignorance.
It seems that the growing number of pupbpeer entries is actually working against the original goals of the site as the large numbers of irrelevant or outright nonsense comments (like a possible gel splicing in a ’97 PNAS paper) are decreasing the visibility of truly worrisome irregularities.
AMW, you strike on an absolutely crude and horrible truth about PPPR. Most, if not all of those currently conducting PPPR are doing so voluntarily. In my case, I now dedicate about 30% of my daily time to PPPR, simply because it is essential and urgent that something be done NOW, while we still have the opportunity to hold authors, editors and publishers accountable, before some of them cease to exist. But the process is painful, there is no glory in anonymity (if anything, we are treated with disrespect and disdain by peers and colleagues for exposing the truth), and we suffer, psychologically and even physically. The unknown war of scientific activists’ struggles may never be known. So, indeed, you are right, when the vast majority of the community (broadly speaking) is actively trying to ignore the criticisms of the literature, when the tiny fraction that gets pointed out actually gets ignored, when authors fail to respond, or when editors or publishers fail to correct the literature, then what “high profile” means are left for us to expose, as you correctly point out, the errors and call attention to all concerned parties, including to fraud? Even if we share disagreements with the main proponents, these are simply individual personal disagreements, absolutely minor and insignificant differences, we are currently limited to three blogs to get our message across: Jeff Beall’s http://www.scholarlyoa.com, Retraction Watch and PubPeer. As you point out, PubPeer is wavering, in terms of policies and in terms of clout and participation, so the tools and our methods to spread the word are limited.
The voice of the whistle-blower and scientific activist is consistently suppressed in the light of the manipulated message. I personally do not have FB, Twitter or other social media accounts, simply because they are a pain in the B. But, from PubPeer, for example, Twitter seems to be an excellent tool for lighting up a debate really quickly. YT is good, but as you say, who has the time to make videos when we are already so stressed doing voluntary PPPR work for the community? In my case now, for almost a decade. The struggle to clear up the literature, and to expose, in the worst case scenarios, the fraud and/or the corruption is not self-serving. It is auto-destructive. But there are critics of PPPR who actually have the gall to call the process a witch hunt, trolling and a series of other terms that fail to recognize the basic importance of exposing these cases and these stories. To the PPPR critics, remember one thing: history has always indicated that when those who are right are banned, it simply serves to fortify their resolve, not diminish it. PPPR, when conducted systematically and thoroughly, may at first seem destructive, or negative. But over time, a pattern begins to develop. And that exposes, case by case, the ugly truth about the current state of science publishing.
How is PubPeer wavering?
The fickle finger of fashion fates may be wavering, short attention span may be at play, but the PubPeer website looks as strong as ever when I check it out.
At least PubPeer remains a site where such issues can be documented and available for future users of scientific literature.
Journals won’t address many of the issues described, Universities take forever to develop reports that may never go public, but at least consumers of scientific publications have a site they can check for issues raised.
Is this embarrassing action by PLOS One and Saad’s university a sign of an approaching science “credit crunch”? How long can these parallel universe activities continue, PubPeer with its public evidence of alleged manipulations and possible misconduct on one side and traditional structures of institutions and journals, used to stifle any inconvenient discussion, on the other? How long till students begin cross-checking their professors on PubPeer and pester them and the universities with insolent questions?
Steven, some problems at PubPeer might never be resolved. At least with respect to anonymity. That is why PubPeer and PubMed Commons are essential complementary sites and not alternatives. Please don’t get me wrong, PubPeer is essential, and useful, but the fact that it’s leadership also remains anonymous, could prove problematic in the long run as the volume of cases increase. On the issue of PubPeer, a Tweet send ~2 hours ago by the anonymous PubPeer management (ACLU.org): “We’re arguing in court tomorrow for the right to remain anonymous online.” With background here:
https://www.aclu.org/blog/technology-and-liberty/academic-freedom-and-right-remain-anonymous-online
If PubPeer loses in court, time for the PPPR proponents to run for the hills.
BB states “It seems that the growing number of pupbpeer entries is actually working against the original goals of the site as the large numbers of irrelevant or outright nonsense comments (like a possible gel splicing in a ’97 PNAS paper) are decreasing the visibility of truly worrisome irregularities.”
1. Splicing is not irrelevant. Even in older papers. Whether it was acceptable in 1992 or 1997, or 2004 is a matter of debate.
2. Each person has a different set of priorities and perspectives. What is irrelevant to you may be relevant to many others, or to the history of science itself. Filtering out comments on perceived manipulation from the outside would be dangerous for PubPeer or for any PPPR site that believes in fair, valid and free (moderated) speech. All valid comments should be respected, even frivolous ones. Because they may carry importance (except to those who are being scrutinized, of course). One cannot exclude potentially “weak” comments based on some random, subjective weighting. All comments carry weight.
3. Unless a journal explicitly stated in its instructions for authors that splicing was ACCEPTABLE, the corrolary is true: splicing was never acceptable. The argument used by some who appear to be quivering is that it was an “acceptable practice”. To those defenders, I state: prove your claim with documentation, and not just with opinions. The world consists of about 190 countries (give or take a few that have come or gone), so make sure that the claims hold true in a geographically-independent manner.
4. What does the PNAS 1997 IFA state about figure manipulation or splicing? Maybe you could check some hard-copies in your uni library?
“[the] article was extensively mentioned in the literature”
Heh. The Piltdown Man fossils were extensively mentioned in the literature, therefore they could not have been fraudulent.
Any response from authors/PLoS Cat?
The post has been updated with their response : )