The Journal of Clinical Investigation has retracted a 2010 article after the researchers acknowledged that the paper contained a substantial amount of manipulated or manufactured data.
How much bad data? Enough to sink seven figures. And the authors said they could not produce raw data in another 19 or so figures and four tables.
But, as the notice states, the researcher said that despite the retraction, they stand by the thrust of their article, “Sphingosine-1-phosphate receptor-2 deficiency leads to inhibition of macrophage proinflammatory activities and atherosclerosis in apoE-deficient mice:”
All authors agree to retract the above article due to multiple use of the same images or manipulation of data in Figures 1A, 2D, 5C, 6B, 6C, and 8A and Supplemental Figure 8E. They are also not able to provide some of the raw data that are used in Figures 2A, 2B, 5, 6, 7C, 8, and 9C, Supplemental Tables 1–4, and Supplemental Figures 2C, 3, 4, 5, 7C, 8A–8C, 8E, 8F, 10A, and 10B. The first author, Fei Wang, has admitted his sole responsibility in altering figures. The authors apologize and deeply regret the impact of this action. However, the authors stand behind data showing that genetic deletion of S1pr2 or pharmacological S1PR2 inhibition alleviates atherosclerosis in Apoe–/– mice fed a high-cholesterol diet.
The study has been cited 16 times, according to Thomson Scientific’s Web of Knowledge.
Now, this doesn’t quite constitute a mega-correction — the article has been retracted, after all — but it’s not too far off. If the authors stand by their last claim, then shouldn’t they resubmit it as a separate, untainted publication?
We tried to ask the JCI about that, but the journal told us that
JCI editors don’t comment on manuscripts within the journal.
We pointed out that, technically, the article was no longer in the journal, but were told that
the editors decline to comment on this matter.
So much for transparency.
With the retraction notice like this, I wonder what would happen if I decided submit a manuscript to The Journal of Clinical Investigation in which I would report that “genetic deletion of S1pr2 or pharmacological S1PR2 inhibition alleviates atherosclerosis in Apoe–/– mice fed a high-cholesterol diet”. Would the Editor reject my manuscript outright on the basis that the authors of the retracted paper made the same claim and they still stand by it? If yes, this would be a clever way to claim priority in any highly competitive area. If no, then any claims made in retracted papers should not be allowed to stand and any attempt to the contrary by the authors should be suppressed.
The question what should happen to results that are part of a retracted publication but which are valid is an interesting one. It has not been answered to complete satisfaction, as far as I can tell.
Republishing a retracted study in a corrected revision has been rarely attempted, from what I can tell, even in cases of honest mistake that undoutedly leave at least some points in the retracted paper undisputed. The centenarian genetics study in science comes to mind.
Republishing parts of a retracted study is probably a very difficult task, which is likely the reason why some authos prefer the mega-corrections. Of course, technically the original, now retracted, paper has not been published. So a resubmission should be viewed by all parties without prejudice. And there certainly cannot be a claim of prior science, so a paper published by a second group after the retraction showing the same results can claim to be the first…….(although depending on the specific circumstances, this would likely be heavily debated). Of course, this discussion is irrelevant for most retractions, as most retractions (especially those openly admitting data and figure manipulations) leave little doubt that the whole thing is not trustworthy.
I don’t fully agree that authors should not be allowed to state what they think still holds up in their retracted paper. The retraction notice should make the whole thing as transparent as possible. Transparency needs to include details on what is wrong but can also include what is ok. After all, the retraction notice does not count as a paper for the authors and everybody knows that it is not peer-reviewed, so everybody takes any additional information in the retraction notice with the appropriate grain of salt. Overall, I appreciate all details offered in a retraction notice.
Curiously there seems to have been an almost identical paper (with almost identical findings) published by an American group
http://www.ncbi.nlm.nih.gov/pubmed/20947824
Since retractions are comparitively rare events – at least on these sort of grounds – one always speculates on the motives that underlie them when they occur.
USA still outranks Japan in terms of influence.
Based on the communiqué from JCI, the first author is believed to be the culprit. I’m a bit skeptical… The other authors did not see the multiple use of the same images ? Did they read the paper ?