‘This actually changes everything’: Altered image in 1999 paper raises potential peril for Stanford president

New findings of altered images in research co-authored by Stanford University president Marc Tessier-Lavigne add to the weight of allegations against him, according to experts on research misconduct.

On Tuesday, Stanford announced an investigation into its president following allegations of altered images in four papers co-authored by Tessier-Lavigne over a seven-year period earlier in his research career. A subsequent analysis by Elisabeth Bik, a scientific integrity expert who specializes in identifying manipulated images, found an additional image that raises concern in a paper published in the journal Cell in 1999, two years earlier than the set of papers originally flagged to Stanford.

Tessier-Lavigne was a senior author, but not the lead author, on the Cell paper. The finding nonetheless points to a pattern of image manipulation on studies that Tessier-Lavigne has worked on across institutions.


Stanford’s investigation was announced following a report in the Stanford Daily that was the first to reveal the misconduct claims. The student paper also reported that the university’s investigation could include the fifth Cell paper.

The newly identified apparent manipulation in Cell is especially serious as it seems to alter the results and appears to be intentional, said Bik. “I would testify in court that’s been digitally altered,” she told STAT. “This actually changes everything. … It’s a more severe level of digital altering.”


The image in question is a western blot, which is used to determine the presence of specific proteins in biological samples. The authors claim the results shown in this western blot help demonstrate that two proteins only form a complex in the presence of a signaling molecule, which in turn, is key to supporting the paper’s findings about how these protein complexes guide the direction of the growth of axons, the long, winding projections neurons send out to connect with other neurons.

Rather than simply improving appearances, as some alterations may do, this one “appears to be changing the actual results,” said Bik.

In this figure, from the 1999 Cell paper, Elisabeth Bik identified parts of a single western blot that appear to have been copied and pasted multiple times. Cell, 1999; annotated by STAT

The level of detail in the alteration also appeared to be intentional, she added. Some image manipulations in other Tessier-Lavigne papers are duplications, meaning the same western blot is used to represent different results, and researchers could conceivably use the same image twice by mistake. “If they had reused the same blot, you could argue negative blots all look the same, we just grabbed the wrong blot,” said Bik. In this figure, though, parts of a single western blot have been copied and appear multiple times.

There are several other instances of apparent alteration in the Cell paper, which have previously been flagged on PubPeer, a site where researchers can discuss potential issues in papers, but are far more minor. The university told Stanford Daily that smaller errors initially described by the student paper in its first piece had been brought to Tessier-Lavigne’s attention in 2015, and that Cell decided a correction was not necessary.

The newly identified image, Bik said, was far more serious than those previously publicly flagged.

Cell did not respond to questions about whether it was previously aware of this particular image. The journal is aware of the concerns, said spokesperson Joseph Caputo, but said the journal’s policy is not to “publicly comment on specific cases or whether an investigation is underway.”

Mike Rossner, president of Image Data Integrity, a consulting firm in San Francisco, and former managing editor at the Journal of Cell Biology, said that the quality of the images meant that he could not say for sure that they were duplicated. But, he wrote in an email to STAT, “If indeed the indicated regions are duplicated, it would take deliberate action using image-processing software to produce those image panels.”

The latest findings raised enough of a concern that Cell should request the original source data from the papers’ authors, Rossner said. Supplying such data would likely be difficult given that the paper is more than 20 years old.

Paul Brookes, professor at the University of Rochester Medical Center in New York who created a blog to flag signs of scientific fraud, noted that it can be hard to determine when duplication is intentional and when it is accidental. ”But in this case I’m inclined to agree that sections of background appearing to be cloned does not seem to be something that could have occurred accidentally,” he wrote in an email to STAT.

In the decades since 1999, the Cell paper has been cited more than 850 times, including in multiple influential reviews published in Science and Nature. It outlined how a molecule called netrin works with multiple protein complexes to attract the axons of neurons to the midline of the brain, and then pushes the axons past it into the opposite lobe from where they originate. This process is essential for coordinating movement on both sides of the body.

The pair of 2001 Science papers now under investigation by Stanford also deal with netrin signaling, demonstrating what other proteins are involved in its ability to attract axons to the midline and how it interacts with other signaling molecules, including one called Slit, to expel neurons past it. These and other publications authored by Tessier-Lavigne helped elucidate the molecular mechanisms that guide developing neurons’ journeys across the brain, decoding the chemical cues that cells use to find their way and form functional circuits.

In this figure from a 2001 study in Science, images of an axon before and after netrin-1 treatment appear to be identical, with only slight changes in cropping that give the impression of growth. Science, 2001; annotated by STAT

Those insights have earned Tessier-Lavigne numerous prestigious awards, including the 2020 Gruber Neuroscience Prize. He has also been a member of the National Academy of Sciences and the National Academy of Medicine since 2005.

The 2003 paper, published in Nature, investigated the role of a different class of signaling molecule, called semaphorins, in guiding developing blood vessels so they can properly transport oxygen to different tissues.

The new findings also raise the stakes for the investigation into Tessier-Lavigne because they appeared two years earlier than the papers originally flagged, which were published between 2001 to 2008. In an earlier statement to Stanford Daily, university spokesperson Dee Mostofi said the images in question did not affect the data or interpretation of the 2001 papers, and that Tessier-Lavigne was not involved in creating the images in the 2003 and 2008 papers.

Valérie Castellani, a professor at the Institut NeuroMyoGéne in Lyon, France, and the first author on the 2008 EMBO paper, told STAT via email that Tessier-Lavigne’s contributions to the study were limited to sharing materials and discussing the scientific approach. His lab was never “involved in the generation and presentation of the data that has been questioned,” Castellani wrote.

Bik said she disagreed with Stanford’s defense. “To me, automatically dismissing it is the wrong answer, it doesn’t show to me they care about scientific integrity,” she said. “At some point, you can’t just blame it on an error, it seems to be a pattern.”

At the time of the 1999 Cell paper, Tessier-Lavigne was a professor at the University of California, San Francisco, and a researcher at Howard Hughes Medical Institute. Some co-authors on the Cell paper also worked with Tessier-Lavigne on two papers published in Science in 2001. When the most recent paper showing seemingly altered images was published in 2008, Tessier-Lavigne was working at biotechnology company Genentech and had completely different collaborators.

“These problems seem to be following him along his career,” said Bik. Even if Tessier-Lavigne didn’t directly create the images, she said, the fact that these issues have occurred on papers he’s involved in alongside multiple teams makes them harder to dismiss as the work of a wayward graduate student.

Tessier-Lavigne was lead author on two of the five papers in question. But all paper authors, including those who didn’t directly create the images, are responsible for the integrity of data, said Rossner. “In my opinion, corresponding authors should review composed figures with the author who prepared the figures by comparing them to the source data to ensure that they are accurate representations of the source data.  Whether failure to do so fulfills the definition of misconduct is a matter of debate, which will have to be decided by any inquiry/investigative committee,” he wrote in an email to STAT.

Research labs are often highly hierarchical, and Bik said that, though she did not know details of Tessier-Lavigne’s lab culture, a senior figure who’s repeatedly associated with manipulated images could be creating an unhealthy work environment. “It might suggest there’s an atmosphere he creates in his labs. Maybe intense pressure or an unwillingness to accept a result giving him different results than he wanted. … That’s not an unusual situation,” she said.

Stanford spokesperson Mostofi said the investigation will be overseen by the university’s board of trustees. “The university will assess the allegations presented in the Stanford Daily, consistent with its normal rigorous approach by which allegations of research misconduct are reviewed and investigated,” she wrote to STAT. Tessier-Lavigne has said he will cooperate. Neither Stanford nor Tessier-Lavigne responded to specific questions about the new findings regarding the Cell paper, nor about whether an outside committee should investigate instead, given that Tessier-Lavigne is a member of the board of trustees.

There is immense conflict of interest in the board investigating their colleague, said Bik. “I have little hope something will come out of it,” she said. “Institutions tend to protect people in power.”

This story has been updated with comments from Valérie Castellani, the first author on the 2008 EMBO paper.

Source: STAT