Postdoc fudged epigenetic data

A former postdoctoral fellow at Washington State University has reportedly falsified data presented in two figures of an epigenetics paper, according to the Office of Research Integrity (ORI) linkurl:report;https://ori.hhs.gov/misconduct/cases/Chang_Hung-Shu.shtml released late last month.

Image:flickr/muhbabyfish

The data fabrication resulted in the retraction of a 2006 __Endocrinology__ linkurl:paper,;https://endo.endojournals.org/cgi/content/abstract/147/12/5524 but a repeat of the original study, which uses newer and more quantitative technology and confirms the paper's conclusions, will be published next week in __PLoS ONE__. "This was an extremely difficult issue for myself and the laboratory to deal with," said linkurl:Michael Skinner,;https://skinner.wsu.edu/ a professor of reproduction and environmental epigenetics at WSU who headed the research. According to the ORI's report, Hung-Shu Chang, a visiting postdoc from Taiwan who worked in Skinner's lab from 2005 to 2006, falsified sequencing data used to identify DNA regions in rat sperm cells that had different methylation patterns following treatment with an endocrine disruptor known as vinclozolin. The same methylated regions (some of which had been falsely identified) were also shown to be present in sperm three generations after vinclozolin exposure -- supporting the hypothesis that changes in the methylation pattern of DNA induced by environmental agents can be passed down to offspring through the paternal allele and contribute to transgenerational disease. "That hypothesis was confirmed with this newer study," Skinner said. "The specific epigenetic sites, though, are completely different." According to Skinner, there weren't any indications of problems during the year Chang worked in his lab. Chang, who had previous experience with the bisulfite sequencing technology used to confirm the DNA methylation sites, regularly presented data during lab meetings. Chang's work came under suspicion, however, soon after he left Skinner's lab toward the end of 2006 to return to Taiwan, when a new postdoc working with newer technology was unable to repeat the analysis of the epigenetic sites Chang had identified. When they turned to Chang's raw data, they found that much of it did not match what had been published, Skinner said. When he contacted Chang in late 2007 about the inconsistencies, Chang had already left science in order to work on his family farm, he said. "Basically then, it was very difficult to get a hold of him." After an additional six months of infrequent communication with Chang, Skinner feared the worst. "I was never really satisfied with his responses," he said. linkurl:Howard Dean Grimes,;https://www.gradsch.wsu.edu/Staff.aspx vice president for research and dean of the graduate school at Washington State University, said that the university investigation was delayed in part because it was hard to get in touch with Chang. "[Chang] had left the United States, left the field of science in its entirety, was somewhere in mainland Asia, and did not have an email or phone contact information that anyone in the US could discover," he said. In late summer of 2008, Skinner alerted the university administration and contacted __Endocrinology__ to issue a retraction of the paper. The journal did not retract the paper until April 2009, however, after the university's research office concluded its investigation. "It's unfortunate it takes that long," said linkurl:Jeffrey Blaustein,;https://www.umass.edu/cns/blaustein/ editor in chief of __Endocrinology__. "but one has to allow for due process to the people who the allegations are made against." Fortunately, the impact of the retraction is likely to be limited, said linkurl:R. Thomas Zoeller,;https://www.bio.umass.edu/mcb/faculty/Zoeller.html an endocrinologist at the University of Massachusetts at Amherst who cited the retracted paper in a linkurl:review;https://endo.endojournals.org/cgi/content/full/147/12/5513 he authored in 2006, also in __Endocrinology__. While the specific epigenetic marks found to be transmitted in the retracted paper turned out to be incorrect, the transgenerational effects of vinclozolin exposure are still well documented in other animal studies. "The retraction of this paper in no way contradicts this original observation," he said. "The worse impact it could have is that people will conclude that the whole field is dirty." "In hindsight, I don't think we could have seen anything," Skinner said. Because it's fairly common to have only one person handling the raw data that's generated in extensive, large-scale experiments, fabrication and falsification of this kind probably often go undetected. "Principal investigators need to be very cognizant of the raw data and the processes by which the raw data was obtained," agreed Grimes. To this end, Skinner has hired permanent bioinformatics personnel to assist the lab members in reviewing complex datasets. "I never before even thought twice about having data analyzed by multiple people to confirm the data is present," he said. "Now I require that two individuals completely go through every piece of data before we publish it." According to the ORI's report, Chang voluntarily accepted the terms of the NIH investigation. __The Scientist__ was unable to successfully locate Chang for comment. H.S. Chang, et al., "Transgenerational epigenetic imprinting of the male germline by endocrine disruptor exposure during gonadal sex determination," Endocrinology, 147:5524-41, 2006
**__Related stories:__***linkurl:10 retractions and counting;https://www.the-scientist.com/blog/display/57449/
[26th May 2010]*linkurl:Authors retract Science paper;https://www.the-scientist.com/news/display/53419/
[26th July 2007]*linkurl:Life After Fraud;https://www.the-scientist.com/2009/07/1/28/1/
[July 2009]