‘I’m nauseated’: Alzheimer’s whistleblower finds possible misconduct by his mentor in their papers | Science

When Vanderbilt University neuroscientist and physician Matthew Schrag went public earlier this year with concerns about apparently doctored images in scores of Alzheimer’s papers—including seminal research underpinning one aspect of the dominant amyloid hypothesis of the disease—he anticipated that his motives and analyses would be dissected. “I also expected every project that I ever participated in to be carefully scrutinized, and that my work would stand up to that scrutiny,” he says.

So Schrag assumed there would be an innocent explanation when, a few weeks after a Science News investigation reported his disturbing findings of apparent misconduct, he received automated emails from PubPeer, a web forum where scientific wrongdoing charges are often leveled. They notified him that two of his own articles from more than 15 years ago had been flagged as containing dubious images.

On close examination, Schrag had to confront an unnerving prospect: that neuropharmacologist Othman Ghribi, Schrag’s first mentor and still a trusted friend, might have also engaged in misconduct.

The papers, published in 2006 when Schrag was an undergraduate working in Ghribi’s lab at the University of North Dakota (UND), covered research on several factors related to amyloid proteins in rabbit brains. (Many Alzheimer’s researchers believe the disease is caused by amyloid’s affects on human brains.) Schrag soon discovered the suspect work in the two papers fit a large pattern of questionable research spanning much of Ghribi’s career, both before and after they worked together.

According to Schrag, in a phone conversation the senior scientist emotionally acknowledged “problems” in many of his papers, including those two co-authored with Schrag, and accepted responsibility. During their discussion, which Schrag recounted to Science, Ghribi maintained to his former student that the underlying findings were correct but admitted to exaggerating data. “I’m nauseated talking about it,” Schrag says.

After initially agreeing to an in-person interview with Science, Ghribi backed out a few days before the appointment, citing a pending UND investigation. He did not respond to a request to verify Schrag’s account of their talk or to comment on a dossier of suspect images in his papers that Schrag has compiled with other forensic experts. But Ghribi told Science via a 5 October email that he wanted to “exonerate [Schrag] of any wrongdoing with the manuscripts he co-authored in my lab.”

Back in August, Schrag contacted a UND official to see whether any of the original images for the two suspect papers could be found. The official said he could not identify any relevant documents. But the university found Schrag’s concerns about the papers, along with PubPeer comments on those and other publications by Ghribi, warranted the inquiry, according to a 10 October notification to Ghribi, obtained by Science.

“We are taking all reasonable steps to secure records related to the research in question,” John Mihelich, a UND vice president, told Science in an email. “We are in communication with the appropriate federal research offices and sponsors.”

Resolving the matter might prove challenging. Ghribi has moved to the University of Texas (UT), Rio Grande Valley. And Schrag says during the recent phone call Ghribi said that, out of remorse, he had discarded his scholarly awards and purged his computers and files of raw experimental data. If so, a review of Ghribi’s papers could be hampered, because proof of image manipulation sometimes requires uncropped original images for comparison with published images.

Schrag recently asked two journals to retract his suspect papers with Ghribi. The publisher of Experimental Neurology said via email that he was awaiting comment from Ghribi. Andrew Lawrence and Marco Prado, editors of the Journal of Neurochemistry, said in an email to Science that they “applied forensic analysis and concluded there are issues with the publication.” They added that all authors, including Ghribi, agreed the paper should be retracted, which will occur soon. Schrag says he’s also evaluating whether images prepared by Ghribi for a third joint paper, which appeared in 2008 in the journal Hippocampus, also warrant a retraction or correction; he has contacted the journal to share his concerns.

After reviewing additional comments posted to PubPeer about other Ghribi papers, Schrag says he recently resolved to take a deeper look at his mentor’s work, including all their joint papers. He enlisted help from microbiologist and forensic image analyst Elisabeth Bik and another image sleuth—a nonscientist who uses the pseudonym Cheshire, in part to minimize legal risks. (Science confirmed Cheshire’s identity and agreed to keep it confidential.) Over recent years, Bik and Cheshire have identified thousands of apparently manipulated or duplicated images in many papers, often leading to retractions or corrections.

In their assessment of Ghribi’s work from 2001 through 2019, the three found suspect images in 33 papers, including the three co-authored with Schrag. Ghribi was the only author shared by all of the papers, and he was usually in a prominent position as either first or last author. The problematic images included Western blots (a common method for displaying proteins in tissue samples) and micrographs of brain tissue. Their joint 67-page dossier, which Schrag provided to Science, shows more than 100 apparently problematic images, many from work funded by grants from the National Institutes of Health. (The agency declined to comment to Science on the matter, but in an email to Schrag, it said it would look into the concerns.)

The three found possible scientific misconduct in five other Ghribi papers in the Journal of Neurochemistry. Lawrence and Prado said those cases “will be also carefully evaluated in a timely way.” Four papers with suspect images, co-authored by Ghribi, appeared in the Journal of Alzheimer’s Disease. UT San Antonio professor George Perry, chief editor of the journal, reviewed the full dossier and says he was surprised to find many images “suspicious,” although he emphasizes that he has no expertise in forensic image analysis. He describes Ghribi as a “likeable,” relatively established researcher whom he has always considered “a person of integrity.”

Perry says he contacted Ghribi to ask for more information on the suspect papers and discuss Ghribi’s role as a member of the journal’s editorial board. The journal will assess the claims, he adds, and issue errata or retractions if warranted.

Schrag acknowledges Ghribi’s early influence in his career and feels sad about his former mentor’s apparent misdeeds. But Schrag says he faced an imperative to correct the scientific record, including work he played a part in. “You have to have a near-religious commitment to research integrity. If the rules apply to others they have to apply to [all of] us,” he says.

This story was supported by the Science Fund for Investigative Reporting.

When Vanderbilt University neuroscientist and physician Matthew Schrag went public earlier this year with concerns about apparently doctored images in scores of Alzheimer’s papers—including seminal research underpinning one aspect of the dominant amyloid hypothesis of the disease—he anticipated that his motives and analyses would be dissected. “I also expected every project that I ever participated in to be carefully scrutinized, and that my work would stand up to that scrutiny,” he says.

So Schrag assumed there would be an innocent explanation when, a few weeks after a Science News investigation reported his disturbing findings of apparent misconduct, he received automated emails from PubPeer, a web forum where scientific wrongdoing charges are often leveled. They notified him that two of his own articles from more than 15 years ago had been flagged as containing dubious images.

On close examination, Schrag had to confront an unnerving prospect: that neuropharmacologist Othman Ghribi, Schrag’s first mentor and still a trusted friend, might have also engaged in misconduct.

The papers, published in 2006 when Schrag was an undergraduate working in Ghribi’s lab at the University of North Dakota (UND), covered research on several factors related to amyloid proteins in rabbit brains. (Many Alzheimer’s researchers believe the disease is caused by amyloid’s affects on human brains.) Schrag soon discovered the suspect work in the two papers fit a large pattern of questionable research spanning much of Ghribi’s career, both before and after they worked together.

According to Schrag, in a phone conversation the senior scientist emotionally acknowledged “problems” in many of his papers, including those two co-authored with Schrag, and accepted responsibility. During their discussion, which Schrag recounted to Science, Ghribi maintained to his former student that the underlying findings were correct but admitted to exaggerating data. “I’m nauseated talking about it,” Schrag says.

After initially agreeing to an in-person interview with Science, Ghribi backed out a few days before the appointment, citing a pending UND investigation. He did not respond to a request to verify Schrag’s account of their talk or to comment on a dossier of suspect images in his papers that Schrag has compiled with other forensic experts. But Ghribi told Science via a 5 October email that he wanted to “exonerate [Schrag] of any wrongdoing with the manuscripts he co-authored in my lab.”

Back in August, Schrag contacted a UND official to see whether any of the original images for the two suspect papers could be found. The official said he could not identify any relevant documents. But the university found Schrag’s concerns about the papers, along with PubPeer comments on those and other publications by Ghribi, warranted the inquiry, according to a 10 October notification to Ghribi, obtained by Science.

“We are taking all reasonable steps to secure records related to the research in question,” John Mihelich, a UND vice president, told Science in an email. “We are in communication with the appropriate federal research offices and sponsors.”

Resolving the matter might prove challenging. Ghribi has moved to the University of Texas (UT), Rio Grande Valley. And Schrag says during the recent phone call Ghribi said that, out of remorse, he had discarded his scholarly awards and purged his computers and files of raw experimental data. If so, a review of Ghribi’s papers could be hampered, because proof of image manipulation sometimes requires uncropped original images for comparison with published images.

Schrag recently asked two journals to retract his suspect papers with Ghribi. The publisher of Experimental Neurology said via email that he was awaiting comment from Ghribi. Andrew Lawrence and Marco Prado, editors of the Journal of Neurochemistry, said in an email to Science that they “applied forensic analysis and concluded there are issues with the publication.” They added that all authors, including Ghribi, agreed the paper should be retracted, which will occur soon. Schrag says he’s also evaluating whether images prepared by Ghribi for a third joint paper, which appeared in 2008 in the journal Hippocampus, also warrant a retraction or correction; he has contacted the journal to share his concerns.

After reviewing additional comments posted to PubPeer about other Ghribi papers, Schrag says he recently resolved to take a deeper look at his mentor’s work, including all their joint papers. He enlisted help from microbiologist and forensic image analyst Elisabeth Bik and another image sleuth—a nonscientist who uses the pseudonym Cheshire, in part to minimize legal risks. (Science confirmed Cheshire’s identity and agreed to keep it confidential.) Over recent years, Bik and Cheshire have identified thousands of apparently manipulated or duplicated images in many papers, often leading to retractions or corrections.

In their assessment of Ghribi’s work from 2001 through 2019, the three found suspect images in 33 papers, including the three co-authored with Schrag. Ghribi was the only author shared by all of the papers, and he was usually in a prominent position as either first or last author. The problematic images included Western blots (a common method for displaying proteins in tissue samples) and micrographs of brain tissue. Their joint 67-page dossier, which Schrag provided to Science, shows more than 100 apparently problematic images, many from work funded by grants from the National Institutes of Health. (The agency declined to comment to Science on the matter, but in an email to Schrag, it said it would look into the concerns.)

The three found possible scientific misconduct in five other Ghribi papers in the Journal of Neurochemistry. Lawrence and Prado said those cases “will be also carefully evaluated in a timely way.” Four papers with suspect images, co-authored by Ghribi, appeared in the Journal of Alzheimer’s Disease. UT San Antonio professor George Perry, chief editor of the journal, reviewed the full dossier and says he was surprised to find many images “suspicious,” although he emphasizes that he has no expertise in forensic image analysis. He describes Ghribi as a “likeable,” relatively established researcher whom he has always considered “a person of integrity.”

Perry says he contacted Ghribi to ask for more information on the suspect papers and discuss Ghribi’s role as a member of the journal’s editorial board. The journal will assess the claims, he adds, and issue errata or retractions if warranted.

Schrag acknowledges Ghribi’s early influence in his career and feels sad about his former mentor’s apparent misdeeds. But Schrag says he faced an imperative to correct the scientific record, including work he played a part in. “You have to have a near-religious commitment to research integrity. If the rules apply to others they have to apply to [all of] us,” he says.

This story was supported by the Science Fund for Investigative Reporting.