MU professor R. Michael Roberts has retracted research published in Science magazine after a nearly yearlong university investigation concluded that accompanying images were doctored by one of his associates, who has apparently fled the country.
Roberts, who was officially cleared of research misconduct by a university committee in February, asked the magazine to retract the research in a letter that will be published in this week’s print issue of Science, due out Friday. The letter appears on the magazine’s Web site today.
In the letter, Roberts, as principal investigator, took responsibility for the doctored images. He said he made the mistake of “placing excessive trust in his co-worker” while failing to assure that “a complete set of raw data existed at the time the questions first arose about the paper.”
In an interview, Roberts said, “I have to take responsibility because it did take place in my laboratory.”
Roberts says he was “overly trusting” of Kaushik Deb, a former post-doctoral researcher who was the lead author of “CDX2 Gene expression and trophectoderm lineage specification in mouse embryos,” which ran in the Feb. 17, 2006, issue of Science.
A recently completed investigation by MU’s Research Responsibility Committee found that Deb committed research fraud by falsifying or fabricating some images of the mouse embryos portrayed in the study. Roberts and two other researchers, Mayandi Sivaguru and H.Y. Yong, were cleared of wrongdoing.
“We had to actually address individual pieces of information and determine whether or not they added up to guilt,” said the committee’s interim chairman, Gordon Christensen, an MU professor of medicine. “There was a whole series of findings we had to assemble and a whole series of conclusions that we had to come to.”
The published research caused an immediate stir in the academic world by presenting evidence that the first two cells of mouse embryos possess markers that indicate from a very early stage whether they will grow into a fetus or placenta. The findings were contrary to traditional beliefs and did not go unchallenged. A scientist outside the university sent a letter to the magazine several months after the article was published questioning the validity of the images of the embryos. Science alerted MU of the objection, triggering the university’s investigation, and later published an editorial expression of concern, alerting readers that the research “may not be reliable.”
Natasha D. Pinol, communications officer for Science, said that of the roughly 950 papers published in the magazine every year, only a “handful lead to some type of clarification or retraction.”
“Unfortunately,” Pinol wrote in an e-mail, “scientific misconduct does occur, though other retractions are due to honest error. The overall number of retractions has been generally constant from year to year, though if the initial report was particularly noteworthy, the retraction is likely to attract more attention as well.”
The committee investigating the fraud allegations was originally chaired by Phil Harter, an MU law professor. Harter resigned in February 2007, asserting that Roberts, Sivaguru and Yong had not been told soon enough that the committee had determined they had no role in the alleged misconduct.
The committee, now headed by Christensen, continued its investigation of Deb, who had aroused suspicion at the outset by abruptly resigning his research position. He then left town, leaving no forwarding address and offering no explanation to superiors.
Robert Hall, MU’s associate vice chancellor for research and director of compliance, said Deb’s failure to appear before the committee “could have been viewed as an admission of guilt” on its own. However, he said MU carried out a full investigation anyway to ensure that the findings would be fair and accurate.
“I really wish he had appeared before the committee,” Hall said. “We had experts on digital imagery from inside and outside the university testify. The committee evaluated all of the evidence. Even Deb got eminently fair treatment.”
Christensen said the committee determined that Deb deliberately altered images of the embryos using photo editing software, including the popular application Photoshop. “The findings were basically that Dr. Deb did indeed fabricate data and the fabrication was very specific to very specific images,” Christensen said.
Although MU’s investigation has concluded, Hall said the committee’s final report could not be made public until a mandatory review by the federal Office of Research Integrity validates the university’s findings; Hall said that could take another year.
Roberts selected Deb to do research on mouse embryos because of his previous experience with them. Although Roberts secured the funding from the National Institutes of Health to perform the research, he said that Deb was working “relatively independently,” and that his own role was limited to that of overseer. Roberts said neither he nor the other collaborators had any indications that Deb was altering the images.
“This was a very hardworking, likeable young man,” Roberts said. “He was quite intellectually stimulating. He appeared to be the absolutely ideal post-doctoral candidate in the laboratory. I thought very highly of him.”
Roberts said he has “reliable evidence” to believe that Deb is in India, having left behind his wife, who was served the charges by the university. The committee’s findings will likely mean Deb will be barred from conducting federally funded research for three to five years.
“He’s done,” Hall said, “He’s been held in front of the international community as being guilty of committing research fraud. This basically is the end of a career.”
Christensen said Roberts was cleared of wrongdoing by the committee, but that there was some concern over “whether had acted appropriately at all times” during the research period. “Since he addressed that in the letter he sent to Science,” Christensen said, “we had no reason to suspect anything other than that he had been tricked.”
For Roberts, the retraction marks the end a lengthy and trying ordeal. Roberts said he continues to work with the mouse embryos, as well as conducting research into the conversion of human embryonic stem cells into placental stem cells. Deb’s fraud, he said, would make him “unfortunately be more skeptical” of his colleagues in the future.
“It’s been a difficult year and a half,” he said. “But I have other research going forward. I still enjoy my work. I have excellent people working for me in my laboratory.”