An article by Genentech scientists received an extensive correction in January for multiple instances of image duplications after comments on PubPeer spurred the authors to review the work.
But the correction “inadvertently omitted” an additional duplication, and will be updated after Retraction Watch brought the matter to the journal’s attention, a representative for the publication said. The sleuth who identified the additional duplication said the original article should have been retracted instead of corrected.
The article, “Death-receptor activation halts clathrin-dependent endocytosis,” appeared in July 2006 in Proceedings of the National Academy of Sciences, with a correction issued that September. It has been cited 99 times, according to Clarivate’s Web of Science.
Most of the authors, including the first and corresponding authors, were affiliated with the biotech company Genentech, based in South San Francisco. When the paper was published, Marc Tessier-Lavigne, the former president of Stanford University, was senior vice president for research drug discovery at the company, according to an online CV. Tessier-Lavigne stepped down from Stanford in 2023 after The Stanford Daily reported on questions about his research, including from his time at Genentech.
In June 2023, an anonymous user on PubPeer commented on elements of figure 2 of the paper that appeared “more similar than expected.” Another user chimed in with an additional finding in the same figure.
Three of the paper’s authors responded in August 2023. The post, signed by first author Cary Austin, second author David Lawrence, and last author Avi Ashkenazi, all Genentech employees, stated the editors of PNAS had alerted them to the query.
“We agree that the images are problematic and are undertaking a comprehensive examination of the image data in the article,” the three authors wrote. “We will then update journal editors to determine an appropriate course of action.”
On January 3 of this year, PNAS published a correction detailing about two dozen instances of splicing, both vertical and horizontal, and duplication in five of the paper’s figures. The authors stated they found the issues with the help of an expert.
“Because the source X-ray films or image files are no longer available, the authors have repeated these experiments,” the notice stated, with reference to a preprint the authors posted on bioRxiv in October 2024. “The new data confirms the original results, reaffirming the experimental conclusions,” they added.
“The editors have evaluated and approved the authors’ response and maintain confidence in the paper,” the notice concluded.
David Sanders, a biologist and scientific sleuth at Purdue University in West Lafayette, Ind., reviewed the original article at our request to look for any issues not noted in the correction. He identified an additional duplication in figure 12, which the correction stated exhibited “potential vertical splicing,” but not duplication.
We reached out to the paper’s authors, PNAS, and Genentech with Sanders’ finding.
Prashant Nair, a representative for the journal, told us the editorial office had “inadvertently omitted” mention of the duplication in figure 12 while finalizing the correction text with the authors. The journal will update the correction notice to include the duplication, Nair said.
The figure “was part of the editor’s original evaluation,” he said. “As stated in the published Correction notice, the editor reviewed the information provided in the authors’ response and remains confident in the findings.”
The authors did not respond to our query. Karen Ring, senior manager for science communications, public affairs and access at Genentech, said the company was “aware of the potential issue you refer to in Figure 12 and are working with PNAS to address it.”
Ring did not respond to our question about whether any other work by the authors was under review. “Genentech is a company founded on principles of scientific excellence and integrity and we remain committed to those principles today,” she said.
Sanders said the original PNAS article “with so many problems” should be retracted. “It simply wasn’t valid.”
The journal has previously allowed authors to correct a paper with reference to data published in a preprint. In August and October of last year, PNAS published two corrections for the same paper in which data had been reused. The authors “performed confirmatory experimentation” posted in a preprint, according to the initial correction.
The practice allows authors “to escape responsibility for the original article,” Sanders said. “I find that an unsatisfactory solution.”
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on X or Bluesky, like us on Facebook, follow us on LinkedIn, 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].