It seemed like a touching tribute when Jiasheng Diao dedicated his 2009 article, “Crystal Structure of Butyrate Kinase 2 from Thermotoga maritima, a Member of the ASKHA Superfamily of Phosphotransferases,” in the Journal of Bacteriology to a deceased mentor, Miriam Hasson.
Before her death in January 2006, of a brain tumor, Hasson and her husband, David Sanders, made up a power-team of protein researchers at Purdue.
Hasson was an X-ray crystallographer while Sanders is a biochemist, and together they would map out the structure and function of proteins. One of their projects was a collaboration funded by a grant to Sanders from the National Science Foundation. By the time Diao joined, the effort had already led to butyrate-kinase crystals — albeit of poor quality, Sanders said.
When Hasson died, Sanders — with his institution’s blessing — took control of her data. That made sense, since their labs had collaborated closely. Indeed, they occasionally shared post-docs, including Diao, who had started with Hasson on a project looking at a protein called butyrate kinase but then moved over to Sanders’ lab as the work evolved.
Diao understandably wanted to turn his labors into papers, but according to Sanders, the trainee was insistent on getting at least two articles out of work the senior scientist believed was worth no more than a single publication:
I wanted to publish this paper, but there were particular points that I wanted to publish. The protein is very similar to acetate kinase. The standard is not to publish all the details of a protein that has a structure that is substantially similar to that of a previously published protein structure; it’s mostly the same protein, except it’s a little different because of two particular features.
Eventually, Diao ended up leaving the lab for a position at the University of Pittsburgh, and never wrote a manuscript. Or so Sanders assumed.
In fact, however, Diao was typing away. The first tipoff came when Sanders found out that Diao had submitted a paper using Purdue data to the Journal of Molecular Biology. He had also included Hasson’s name on the manuscript:
He put her name on it, plus the name of people who had not agreed to authorship nor to having their data included in the article.
Sanders said he contacted the journal and explained the situation, which was sufficient to quash the article.
At the time, Sanders said, the editors requested that Sanders launch an investigation into Diao’s actions. But the researcher demurred:
In my foolishness I said, ‘I don’t want to punish this guy, he made a mistake,’ and he agreed to withdraw [the paper].
Sanders even went a step further, he said, offering to write the paper with Diao — on the condition that it be a single article, not multiple publications. The young man agreed, and Sanders went about preparing a new paper.
He was astonished, he said, when Diao began calling him to demand again that they break the work into multiple manuscripts.
At that point, Sanders said, he gave up on Diao. But that didn’t last long. In 2009, he saw the article in the Journal of Bacteriology, which went as far as to credit Hasson. He immediately contacted the editors, but was told that they were unable to retract the paper before Purdue conducted an investigation into the matter. Sanders brought the case to the attention of university officials, who completed their inquiry sometime last year.
Armed with the results — which, he said, conclusively showed that Diao had used his and Hasson’s data without authorization and committed other acts of publishing misconduct — Sanders again went to the journal and requested the retraction. And finally, more than three years after the article first appeared, comes this notice:
Purdue University has stated that Dr. David Sanders is the sole custodian of the data upon which this publication was based. Since this article was submitted to the Journal of Bacteriology without the prior permission of Dr. Sanders, the publisher hereby retracts this publication.
In fact, the paper in the Journal of Bacteriology — which has been cited twice, according to Thomson Scientific’s Web of Knowledge — wasn’t merely the second one Diao had published — and lost — for misusing Hasson’s data and name. In September 2009, Diao submitted a paper to Proteins on butyrate kinase on which Hasson’s name also appeared, along with a third researcher, Yunglin Ma. The article appeared online in October 2010, and immediately caught Sanders’ attention.
Sanders contacted the journal, which, he said, agreed to retract the article before it went into print. The notice is virtually identical to that in the Journal of Bacteriology:
The following article from Proteins: Structure, Function, and Bioinformatics, “Open and closed conformations reveal induced fit movements in butyrate kinase 2 activation,” by Jiasheng Diao, Yunglin D. Ma, and Miriam S. Hasson, published online on 21 October 2010 in Wiley Online Library (onlinelibrary.wiley.com), has been retracted by agreement between the journal Editor in Chief, Bertrand Garcia-Moreno, and Wiley Periodicals. The retraction has been agreed because it was established by internal investigation performed by Purdue University that the authors of this article are not the owners of the data and have no right to publication.
We do observe that the Proteins notice inadvertently — and erroneously — lumps Hasson with Diao. (We’ve yet to figure out how Ma, a researcher at Washington University in St. Louis, fits in here). Were Hasson still alive, she certainly would have had a “right to publication.” Indeed, it seems safe to assume that Diao chose to use Hasson as a co-author on his papers precisely because she would not be able to challenge his claim on the data.
We tried to reach the editor of the Journal of Bacteriology to find out why it took so long for them to retract the 2009 paper. But a representative of the American Society of Microbiology, which publishes the journal, told us that its editors did not discuss the reasons for retractions as a matter of policy. That strikes us as not quite true, given the accessibility of Ferric Fang, editor of another ASM title, Infection & Immunity, in the Naoki Mori case and others. But pointing that out didn’t produce a response.
After a stint in the Baylor lab of pharmacologist Yongcheng Song, Diao is now at the University of Delaware. When we reached him in the lab, Diao told us that he had grown frustrated with the “delays” in publishing while at Purdue and that he felt he had a right to write papers with the data he’d worked with:
My career already in danger by these delays.
He also said that Hasson had typed her name on the Journal of Bacteriology manuscript before her death. However, Sanders dismissed that claim, noting that he, Hasson, Diao and another co-author had published a 2003 paper in Acta Crystallographica Section D: Biological Crystallography:
We were working on a manuscript for the structures while she was alive with both our names on it not just those of Drs. Diao and Hasson. She would never have consented to his publishing without me or indeed publishing what he published. Of course, her name was going to be on the paper I wished to publish.
It’s a pity that the Journal of Bacteriology and the ASM don’t want to be more talkative about the Diao case, because they might be able to convince another recalcitrant editor to take a closer look at another Diao paper under a similar cloud.
In 2005, Diao left Purdue for the University of Pittsburgh and the crystallography lab of Joanne Yeh. Yeh, a structural biologist recently recruited from Brown University, said she quickly realized that Diao wasn’t a good fit for her federally-funded projects. For starters, she said, he seemed “paranoid” about the potential hazards of x-rays — an odd concern for an x-ray crystallographer — and did not want to work on the tasks she assigned him. So she declined to renew his contract for a second year.
Some time later, Yeh said, she received an email from the editor of a crystallography journal who’d received a manuscript that looked like it came from her lab but hadn’t. The sole author: Jiasheng Diao.
I told the editor that [Diao] had no permission to publish this, she said.
More than a bit irked, she reached Diao and told him that not only had he failed to include his collaborators on the paper, but that the manuscript itself was riddled with errors.
I said, you will not be able to get this published without correcting the space group and other errors, issues that I had pointed out to Diao when he was still in my lab but to no avail, as he insisted that he was correct. I also stressed to Diao that before the paper is even submitted, you need to be sure that everyone gets to read the draft you intend to submit and give their permission to have their names on it.
Diao agreed, Yeh said, and she sent him a disk with the original data so that he could reprocess the data and correct the model and interpretations.
Then, in January of this year, Yeh was doing an online lit search when she stumbled upon a paper in Protein Science, “Crystal structure of a super leucine zipper, an extended two‐stranded super long coiled coil,” listing Diao as the sole author, although the acknowledgments section does contain the following lines:
This work was supported by the AFSOR [sic – that should be Air Force Office of Scientific Research, or AFOSR] grant F49620-03-1-0365 to Professor Joanne I. Yeh. The author thanks Professor Joanne I. Yeh for permitting me to publish this work. The author thanks Antoni Tortajada for crystallizing the leucine zipper mutant peptide.
Yeh contacted the journal’s editor, Brian Matthews, and the two went back and forth about an erratum or a letter to the editor to correct the record. Yeh felt the paper was still deeply flawed.
I even went so far as to independently reprocess the data, solve, refine and deposit the coordinates and structure factors in the protein databank as soon as I learned about this in 2010 so that correct structure and data would be available to the public.
But Matthews eventually grew skeptical of Yeh’s claims, and decided that he would no longer print either an erratum or a letter from her. As for the dispute over authorship and rights to the data, Matthews said, he relied on Diao’s word over Yeh’s:
In his article, he has in his acknowledgment a statement that he had permission to publish the work. He has provided to me documentation that that is a correct statement.
We asked Matthews whether, in light of the recent retractions of Diao’s work and the Purdue investigation — which seemed to satisfy the Journal of Bacteriology — he might be willing to revisit the case. But he demurred.
Yeh, for her part, said she’s ready to move on — but less out of a sense of closure than of fatigue.
I spent a good six months of my life, along with some of my folks, to get this thing corrected. Protein Science dropped the ball. I want to at least acknowledge those who contributed to a great majority of the work – Antonio Tortajada who grew the early crystals at Brown, and some of my current group members, Shoucheng Du who grew and optimized the crystals, Regina Kettering and John Jeff Alvarado both of whom I pulled in to help me to quickly but correctly solve and refine and deposit the SF and coordinates in 2010 so that a valid structure of the peptide (accession number 3M48) would be publicly available from the RCSB [protein] databank and the BNL synchrotron source where I collected most of the peptide data.
Update 6/28/12: We’ve changed the headline on this post after a lengthy exchange with Dr. Diao, who objected to the use of the word “thief.” We thank him for his input.