US postdoc fabricates DNA data
A former postdoc at the University of Nebraska Medical Center (UNMC) falsified and fabricated DNA sequences and methylation status in unpublished data about a tumor suppressor gene, a UNMC investigation, in conjunction with the Office of Research Integrity (ORI), has found.
From 2002-2005, Lois Bartsch worked in James Shull's laboratory at UNMC, researching the tumor suppressor gene, p16Cdkn2a, in rats. The investigation concluded that Bartsch altered the nucleotide sequence of the p16Cdkn2a pr
A former postdoc at the University of Nebraska Medical Center (UNMC) falsified and fabricated DNA sequences and methylation status in unpublished data about a tumor suppressor gene, a UNMC investigation, in conjunction with the Office of Research Integrity (ORI), has found
From 2002-2005, Lois Bartsch worked in James Shull's
laboratory at UNMC, researching the tumor suppressor gene, p16Cdkn2a
, in rats. The investigation concluded that Bartsch altered the nucleotide sequence of the p16Cdkn2a
promoter region in GenBank, in a National Cancer Institute grant application, and a poster presentation.
The investigation also found that Bartsch's grant application contained falsified data
on the methylation status of rat p16Cdkn2a
, and a fabricated claim of a polymorphism in the human homologue. No data was ever published in any peer-reviewed journal.
A two-year inquiry was initiated after Shull asked the university to investigate around 2005, David Crouse
, UNMC's associate vice-chancellor for academic affairs, told The Scientist
. Shull declined to comment in any detail, but said that his suspicions of fraud led him to ask a second person in his lab to repeat some of Bartsch's experiments, and her data could not be replicated.
"These were pretty distinct findings that couldn't be rebutted," Crouse said. "The data simply had been altered."
According to the ORI's report, Bartsch entered into a "Voluntary Exclusion Agreement" in which she neither admits nor denies the misconduct findings. Bartsch
, now an assistant professor at Graceland University in Lamoni, Iowa, where she teaches but does not maintain an active laboratory, did not respond to requests for comment.
Linda Buckles, a former PhD student with Shull who switched supervisors and continues to work as a postdoc in another UNMC lab, sequenced the p16Cdkn2a
coding region in Shull's lab before Bartsch starting working on the promoter. "My data conflicted with her data," she said. Disagreements between Buckles, Shull, and Bartsch eventually forced Buckles to leave the lab non-voluntarily in February 2004, she said. "Dr. Shull and I will never be on speaking terms or have a good relationship," Buckles said.
Karen Deffenbacher, a former postdoc who joined Shull's group in 2004 and now works in a different UNMC lab, said that problems in Shull's lab went beyond Bartsch's data. "There was a general problem of integrity of the work coming out of that lab," she said. "[Shull] didn't really audit any of the data. He just wanted answers that matched his hypothesis."
In 2006, Deffenbacher said Shull asked her to be the principal investigator on a Department of Defense grant
originally held by Bartsch. This funded her research for a year, after which she wrote a final summary report. This was no easy task, she said. "It was a mess," because a number of different people had all worked on the project. "It wasn't clear where all the data was."
Both Buckles and Deffenbacher told The Scientist
that they have draft manuscripts that Shull refuses to publish. Shull, however, said that both former lab members had their papers rejected from peer-reviewed journals, and that he is currently incorporating their results into other manuscripts, or re-analyzing portions of the data. Deffenbacher said that only one of her two manuscripts had been rejected though, and that Shull was sitting on both of them without her consultation.
Other than Bartsch's misconduct, Crouse said he didn't know about any other problems in Shull's lab.