There is a “high possibility” that fraud was involved in a series of RNA studies from the laboratory of biochemist Kazunari Taira, a year-long investigation by the University of Tokyo has found. The papers in question “had no reproducibility and no credibility”, according to a statement dated 29 March.

The committee that carried out the inquiry also claims that a researcher on Taira's team may have faked data for the investigators, hoping to convince them that the results were reproducible. “Raw data submitted to the committee were fake,” says its chair, Yoichiro Matsumoto.

Co-author Hiroaki Kawasaki allegedly printed data for the committee obtained using software that was not available in 2003, when the experiment was conducted.

Kawasaki told Nature that he had possessed a demonstration copy of the software. The software maker, Applied Biosystems, told both the committee and Nature that no demonstration copy was available at the time.

Tipped off by a whistleblower in Taira's lab, the committee also tracked down a record that Kawasaki bought an enzyme he claimed to have constructed himself. Kawasaki says he bought it as a control when repeating his experiment. “I've never done anything wrong,” he says.

Taira has asked for retractions of four papers in question; he had already requested retraction of an earlier Nature paper1. So far, Kawasaki has not agreed to sign off on the retractions.

Taira's laboratory has developed novel RNA technologies to identify genes related to cancer and other diseases, preventing them from being expressed. In one paper2, he and his colleagues reported that a human enzyme was expressed in the bacterium Escherichia coli, and this could be useful in making small fragments of RNA to suppress genes.

A second paper3 explained that introduction of small RNAs can induce DNA methylation — a process involved in some diseases such as cancer — in mammalian cells. The phenomenon had been reported in plant cells. The paper was later corrected, but not retracted.

On 26 March, Taira requested retraction of these two studies, along with two others4,5. In his statement to the committee, Taira says he requested the retractions because of “a problem of research ethics”, and that Kawasaki had failed to provide lab notebooks to support the original results. Nature is deliberating on the retraction request, says editor-in-chief Philip Campbell.

When contacted by Nature, Taira said he stood by everything in his statement. He has in the past denied any wrongdoing.

A University of Tokyo investigation has found the possibility of wrongdoing by some of its researchers. Credit: JTB PHOTO/ALAMY

Questions were first raised in April 2005 when the RNA Society of Japan, noting complaints from researchers, asked the University of Tokyo's School of Engineering to investigate whether 12 of Taira's papers were reproducible. The university selected a few of the papers and asked Taira to submit raw data by September. He failed to do so, blaming Kawasaki for allegedly not keeping good notebooks and storing back-up data on a computer that broke. The committee extended the deadline to the end of March, but again no evidence of reproducibility emerged.

A separate investigation was undertaken by the National Institute of Advanced Industrial Science and Technology (AIST), where Taira had headed a gene-discovery centre. Misconduct in nine of the twelve papers, it reported on 3 March, “could not be ruled out”. One other paper was in the clear, and two were not scrutinized as they did not involve AIST research. Last week, the institute announced that it would not renew Taira's contract, which ended on 31 March.

A separate committee, set up earlier this year at the University of Tokyo, interviewed Taira, Kawasaki and other lab members about the lab's management and the possibility of misconduct. It plans to report to the university's president by 10 April, and to pass the issue to the university's disciplinary panel, which will decide possible penalties in the next few months.