A group of neuroscientists in Switzerland have retracted a 2019 paper in Science whose first author they say falsified data in the study.
The article, “Insular cortex processes aversive somatosensory information and is crucial for threat learning,” came from the lab of Ralf Schneggenburger, of the Ecole Polytechniqe Federale De Lausanne (EPFL). The first author was Emmanuelle Berret, then a post-doc in the lab.
EPFL issued a press release about the study when it appeared. According to the release, the research showed that the insular cortex — a region “deep within the lateral sulcus” — is in charge of processing how mice and humans (pace, James Heathers) apparently learn from painful stimuli:
By switching off the insular cortex during the painful event, the scientists found that mice became essentially fearless against a mild electric shock to the foot. In addition, the ability of the mice to learn from the painful event was greatly reduced.
The study demonstrates that, besides informing our brain about bodily states, the insular cortex can send a strong warning signal to other brain areas involved in forming a memory of the unpleasant event. ‘Because silencing the insular cortex takes away the unpleasant feeling normally associated with a painful event, our study suggests that neurons in the insular cortex cause the subjective feeling of pain, and induce learning about the pain in other brain areas,’ says Schneggenburger.
Shortly after publication, a poster called “Lataxiena Fimbriata” noted on PubPeer that:
The videos in the supplementary information, in particular video 3, appear as though opto stimulation is producing behavior induced seizure activity that the authors describe as “aversive behavior”.
The comment prompted this reply within days by Ralf Schneggenburger, the corresponding author of the paper:
Dear “Lataxenia” thank you for sharing this important observation with us. If you would like to discuss it more, please reveal your name, or write me an e-mail. best regards, R.S.
The Swiss group reanalyzed their data in October. Schneggenburger told Retraction Watch that
The re-analysis of the published data was prompted by a new analysis method that we had meanwhile developed in the lab (higher time resolution). We simply wanted to see how the published data would look like when it’s analyzed at higher time resolution. It was then when we found the difference…
According to the retraction notice:
The May 2019 Research Article, “Insular cortex processes aversive somatosensory information and is crucial for threat learning,” used optogenetic methods in mice to conclude that insular cortex is involved in auditory cued fear learning. A reanalysis of the data performed by the authors in October 2019 showed, however, that the mouse behavior data reported in Figs. 1C, 3C, 3F, and 6B, and the corresponding data in supplementary figures, had been manipulated. The reanalysis showed that data points from many individual mice had been moved, with the effect that the difference between optogenetic silencing groups and control groups became larger than in the real data. Thus, in the reanalyzed data, the statistical significance disappears for many datasets of Figs. 1, 3, and 6, and these experiments need to be reestablished in future work. The first author, who performed these measurements, has admitted to having committed the data falsification. No other coauthors were involved in the data manipulation, and thus their data remain valid. Because the data manipulations affect important conclusions of the paper, the authors retract the Research Article and apologize to the readership of Science.
Schneggenburger told Retraction Watch:
Importantly, no other co-authors were involved in the data manipulations, and their data in the paper remain valid. Nevertheless, because the data manipulations affect major conclusions of the paper, we decided, together with the editors, to retract the paper. The first author, who made the data manipulations, has left EPFL with effect of Nov. 15th. There is no further investigation from the side of the EPFL in this case. The responsible persons at EPFL decided that such an investigation was not necessary, because we (the co-authors) immediately made a self-report of the data manipulations within EPFL as well to Science (on Nov. 1st), and because we retracted the paper.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, 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].