Although an investigation released today exonerates him of research misconduct, Stanford University President Marc Tessier-Lavigne will step down next month after the probe also concluded that at least four of his papers dating back as far as 1999 contained data manipulated by a member of his lab.
The probe, commissioned by Stanford’s Board of Trustees, says the prominent neuroscientist “failed to decisively and forthrightly correct mistakes in the scientific record” when concerns arose about his papers. The report further faults him for creating a lab culture that fostered an unusually high rate of data integrity problems.
The report “clearly refutes the allegations of fraud and misconduct that were made against me,” including allegations of a cover-up around when he was at the company Genentech, Tessier-Lavigne said in a statement also released today. But the award-winning researcher, whose work ranged from spinal cord development to Alzheimer’s disease, also acknowledged that the investigatory panel “identified some areas where I should have done better.” Tessier-Lavigne added that “for the good of the university” and in anticipation of “debate about my ability to lead,” he would end his university presidency on 31 August.
Stanford’s board said it had accepted his resignation and “agrees with him that it is in the University’s best interests.” Richard Saller, a professor of European studies, will serve as interim president. Tessier-Lavigne will remain a faculty member and will continue to run his lab.
Elisabeth Bik, a consultant on image manipulation who first flagged some of the problems in his papers, says it was appropriate for Tessier-Lavigne to step down as Stanford’s leader. “There were multiple cases of misconduct done under his watch. … I feel his role as the president of the leading research university in the U.S. is no longer trustworthy.”
Some Stanford observers weren’t surprised by the resignation. “I doubt that anybody is going to be shocked. This has been a cloud hanging over Stanford for quite a while now. And it was pretty clear that it was a serious cloud,” says Hank Greely, a Stanford law professor and bioethics scholar. “This is enough exoneration that he can go out with dignity. … It’s probably better for the university to get a new leader. And I think Marc is certainly smart enough to realize that he’s not indispensable, nobody’s indispensable.”
The developments today nevertheless shook some in the neuroscience community. “In my experience, Marc is a person of the highest integrity, intelligence, and generosity,” says Cori Bargmann, a neurobiologist at Rockefeller University who describes Tessier-Lavigne as a longtime friend and colleague. “I’m sad that he made the decision to step down from the presidency of Stanford, because he was excellent in that role as well.”
Tessier-Lavigne is best known for his work in the 1990s discovering netrins, proteins that guide the growth of nerve cell projections known as axons in the developing spinal cord. His resignation brings an end to an extraordinary 8 months that began in late November 2022, when Stanford’s student paper, The Stanford Daily, reported that The EMBO Journalwas investigating possible image manipulation in a 2008 paper and that questions had been raised about three other articles on which he was a co-author. The image issues had come to the newspaper’s attention via a discussion of them on PubPeer, a forum where scientists, often anonymously, discuss irregularities in papers. Stanford’s Board of Trustees announced it was forming a special committee to investigate.
Concerns about the other three papers—one in Cell in 1999 and two in Science in 2001—had also previously come up on PubPeer in 2015 when Tessier-Lavigne, then president of Rockefeller, was under consideration for the Stanford presidency. At the time, he submitted corrections to both journals, but Science failed to publish them because of an editorial error and Cell did not find a correction was needed. After the new publicity about the papers emerged, both journals in December 2022 added expressions of concern and said they were awaiting findings from the Stanford investigation. Tessier-Lavigne responded that he took responsibility for any errors and said: “The integrity of my work is of paramount importance to me, and I take any concerns that are expressed very seriously.”
In January, the special committee contracted Mark Filip, a former federal judge, and his law firm Kirkland & Ellis, to lead the probe with a panel of five prominent scientists and scientific leaders as advisers. Its review grew to include 12 papers on which Tessier-Lavigne was either a corresponding author or a middle author.
Tessier-Lavigne’s position became more precarious a month later, when The Stanford Dailyreported that while he was an executive vice president at Genentech’s research arm, the company found scientific fraud in a high-profile 2009 Nature paper—it had reported an unexpected role for a protein in causing the neurodegeneration of Alzheimer’s. The Stanford Daily reported, based on interviews with several former Genentech employees, that Tessier-Lavigne suppressed the misconduct findings and refused to retract the paper. The Stanford president called the paper’s reporting “replete with falsehoods” and the allegations “breathtakingly outrageous.” He also noted he later published papers revising the findings and said this was part of the normal process of science.
In April, Genentech released its own investigation noting that although its scientists at the time could not replicate the Nature papers’ findings, the company found no evidence its officials had looked into possible misconduct in the work. But the company also pointed to a problem with other research from Tessier-Lavigne’s lab. A year earlier it had dismissed a postdoc in his lab for research misconduct on a different manuscript that was withdrawn before publication.
Today, the Stanford board’s special committee released the law firm’s and scientific panels’ findings, which are based on more than 50,000 documents, interviews with over 50 people, and input from forensic science experts. Its report finds that for seven papers on which Tessier-Lavigne was a middle, or secondary, author, he bears no responsibility for any data manipulation. The primary authors have taken responsibility and in many cases are issuing corrections.
But the 22-page report (plus appendices) found “serious flaws” in all five papers on which Tessier-Lavigne is corresponding or senior author: the 1999 Cell paper, the two 2001 Science papers, a 2004 Nature paper, and the 2009 Nature paper from Genentech. In four of these studies, the investigation found “apparent manipulation of research data by others.” For example, in one case, a single blot from the 2009 Cell paper was used in three different experiments, and a blot from that paper was reused in one of the 2001 Science papers.
The 2004 Nature paper also contains manipulated images, the report found. Although the report says the allegations of fraud and a cover-up at Genentech involving the 2009 Naturepaper were “mistaken”—people likely conflated the fraudulent paper a year earlier, and Genentech scientists’ problems replicating the work, it suggests—that paper showed “a lack of rigor” that falls below standards.
The report finds that overall Tessier-Lavigne “did not have actual knowledge of any manipulation of research data” and “was not reckless in failing to identify” the problems in the papers. Yet it concludes that he did not respond adequately when concerns were raised about the papers on PubPeer or by a colleague at four different points over 2 decades—most recently in March 2021. For example, it chides him for failing to follow up when Science did not publish the corrections he submitted.
The report also faults Tessier-Lavigne for his “suboptimal” decision not to correct or retract the 2009 Nature paper, despite “vigorous discussions” about what to do; instead, he and colleagues published follow-up papers revising the findings. Without “an appropriate appetite” for corrections, “the often-claimed self-correcting nature of the scientific process will not occur,” the report says.
The investigation laid some of the blame for the “unusual frequency of manipulation of research data and/or substandard scientific practices” on the culture of Tessier-Lavigne’s labs. Although several former postdocs spoke positively of Tessier-Lavigne’s mentoring, some interviewees described a culture that rewarded “winners,” postdocs who produced favorable results, and marginalized “losers.” The report found no evidence that Tessier-Lavigne had been in favor of this dynamic, but suggested there were missed opportunities to “improve laboratory oversight and management.”
Tessier-Lavigne says he expects to retract the Cell and two Science papers and correct the two Nature papers. Holden Thorp, Science’s editor-in-chief, says the retractions are already in motion. (Nature has not yet responded to a request for comment.)
Despite the findings of data manipulation, the report does not assign responsibility to any specific members of Tessier-Lavigne’s lab or determine whether the data manipulation fit the federal definition of research misconduct, “fabrication, falsification, fabrication, or plagiarism.” Whether the findings should be reported to the federal Office of Research Integrity will be up to Stanford, Filip says.
The report says Tessier-Lavigne should have not been expected to catch all the data manipulation before publication, although lab heads typically are considered responsible for what those under them produce. In his statement and a linked document, Tessier-Lavigne said the review revealed “new information” indicating that a specific scientist “engaged in serious data manipulation” in the 1999 Cell paper and the two 2001 papers. He also points to data manipulation “by another individual” in the 2004 Nature paper.
Bik, who spoke with the investigatory panel, says the probe “was done very thoroughly” and calls Tessier-Lavigne’s decision to retract or correct the papers “reasonable.” But she disagrees with the report’s finding that Tessier-Lavigne should not be blamed for not catching the data problems in his papers. “Those are early-career scientists and as a principal investigator, your role is to supervise and mentor and, and oversee the integrity of the data in those labs. So I feel he is responsible … for those results.”
Bik, Thorp, and others suggest the episode raises questions about whether scientists can properly oversee large, active labs when they also have major, administrative positions. Bik further questions whether Tessier-Lavigne should still be leading a lab given the past problems.
For his part, Tessier-Lavigne says he “will be further tightening controls” in his current lab at Stanford, for example by matching processed images with raw data. “I will vigorously apply this and other best practices to ensure that these kind of problems do not recur,” he wrote.
