Last month, the Journal of Bacteriology retracted a structural biology paper that was submitted by postdoc Jiasheng Diao without the knowledge or consent of his former principal investigator (PI) David Sanders, who supervised him on the project. In this article, Sanders shares his account of the story as well as what he has learned from the experience regarding authorship, communication, and the postdoc/PI relationship.
Sanders: It goes back to the fact that there was a long-standing collaboration between Dr. Miriam Hasson, who was my wife, and myself working on phosphotransferases. Together, we determined the structure of acetate kinase. In fact, a graduate student in my lab did that determination. We were working on this project together for a long time, and each of us had a grant to work on the project. She had a National Institutes of Health grant, and I had a National Science Foundation (NSF) grant. Some people in my lab were working on the project, and some in hers, so we had joint group meetings. We really supervised the project together.
To pursue the project further, we decide to study butyrate kinase. Acetate is a two-carbon carboxylic acid, and butyrate is four-carbon carboxylic acid. We were interested in how the enzyme evolved to use four carbons instead of just two and had certain ideas about how that would occur and what the structural changes could be. So, we proceeded to clone, express, and purify the protein. We started crystallization studies.
During this time, my wife had a brain tumor. There were certain times when I had to take full control of the project because of her treatments. Everyone in the lab knew the arrangement. In fact, her specialty was on Vitamin B1 enzymes and mine was phosphotransferases. Everyone understood that was the way it worked.
Then, Diao joined Dr. Hasson’s lab as a postdoctoral fellow. We already had crystals. Eventually Diao got diffracting crystals of butyrate kinase, and we published a paper on the crystallization. He had two different structures of the butyrate kinase, and we submitted the structures to the Protein Data Bank (PDB).
Then, Diao was done in Dr. Hasson’s lab. He had not yet written up a paper, but we were planning to do that. And for these collaborations, I was in charge of writing the papers because it was my field of expertise.
In the interim, Diao found another job on campus with a different advisor, but was there only a short time.
Diao worked methodically, but he was a very stubborn individual. He wanted to do things his way and no other way. We had many conversations about the interpretation of the structure. I had certain ideas, and he would insist on other things against the evidence.
So, then Diao worked in my lab. I paid for him on my NSF grant. His full-time occupation while he was working for me was to write the paper. And he didn’t do that. He made little progress before he left. This was all before Dr. Hasson died.
BTN: During that time that he was in between these labs, it was clear to everyone that you and your wife were on the same projects, and when your wife was unavailable, that you were in control of the data.
Sanders: That’s right. We worked very closely. There was no question in anybody’s mind. Even on her other project, the vitamin B1 project, her collaborators would frequently come to consult with me. I helped her take care of things. She wasn’t just my collaborator, she was my wife. When she died, the vitamin B1 phosphate project that she was working on was also incomplete. That project involved other people, and those people came to me to ask me to look at their papers before they submitted them for publication. They knew she consulted with me on the project. I approved their publications, and tried to make sure everything worked out.
BTN: Were these postdocs in her lab or outside collaborators?
Sander: It involved a postdoc in her lab and outside collaborators.
BTN: And they came to discuss a paper that they were publishing?
Sanders: Absolutely. I didn’t demand anything. I didn’t put my name on the papers. But everyone knew the situation, and that this was the appropriate approach.
With Diao, I urged him to write a paper. He had two different structures that were very closely related. I wanted to publish a single high-quality paper using both of the structures. This is standard in the field nowadays. In the 1970s, a structure was a paper. You had a new structure, you published a new paper. But now, there are multiple structures, structures of mutants, of the protein with a substrate, of the protein with an inhibitor. If you have more than one structure, you don’t publish them sequentially. They are published at the same time because they make a better story. Although there was never a butyrate kinase structure published, it was very similar to acetate kinase. You don’t review the whole structure of the protein when it looks exactly the same except for one distinguishing part. And we had this second structure that was interesting.
But I didn’t hear from him for a while. Then I found out that he had submitted an article to the Journal of Molecular Biology. When we were working on the paper, we had an understanding that other people from the lab that actually did the enzymology would be included. They determined the substrate specificity, which brought up some interesting aspects. The paper would have included the enzymology and structural biology.
There are people that believe that every datum they collect is a separate paper, but I believe in an integrated approach. Everyone can see that in my previously published papers. We approach problems from multiple aspects, and publish high-quality papers from there.
BTN: The papers that Diao submitted was just his work, or did it include work from other people in the lab?
Sanders: With the Journal of Molecular Biology paper, he included data from other people in our labs and put their names on that paper, but they didn’t agree to that. So, he submitted data without the agreement of the other authors, which is absolutely a violation of scientific publications.
BTN: And were these other authors were planning to use this data in future publications?
Sanders: Yes, but future publications of that data would have been impossible if Diao’s paper was published in the Journal of Molecular Biology. They were all shocked. They knew that this was my project, and there was never a question.
The conflict was that I wanted to publish one high-quality paper that focused on the novel aspects of the structure, but he wanted to publish two less impressive papers that recapitulated everything that had already been said. We had a basic dispute.
When I learned about the Journal of Molecular Biology, I explained the situation to the editors. I got to see the reviews. Of course, one of the reviewers said that they saw the other structure of butyrate kinase in the PDB and that it would be so much better if the paper would include both structures. Diao accused me of being that reviewer, as if I would review a paper that had my data and the data from people I was working with and wouldn’t have pointed that out in the review.
BTN: The structures have been published in the Protein Database, so this isn’t a case of you sitting on any data, right?
Sanders: That’s right. They have already been put into all the different databases for structural superfamilies. We’re not sitting on data. If you look at the PDB, you will see Diao’s name and my name is on it and Dr. Hasson’s name as well as the person who originally got the crystals. The data are out there, absolutely.
BTN: What reason did Diao give to support his argument for two papers?
Sanders: He just kept repeating: “I need two papers. For my career, I need two papers.”
BTN: So, he believed that the more papers that he published, the quicker his career would advance?
BTN: Do most postdocs share a similar perspective on quantity of papers vs. quality of papers?
Sanders: There are both approaches, and it’s not just postdocs. Some PIs have that same philosophy as well. Some favor high-quality publications, and others favor multiple publications. It’s quality vs. quantity. The reward structure in science sometimes favors one over the other.
But ultimately, I understand the standards I wish to maintain in my publications. The structure is just the beginning, the interpretation is very important. So, that is my expertise. I know how to interpret structure. It would have been best if we united the two structures. It would have been a disservice to the field and a disservice to us to separate that discussion in order to get a second publication. Maybe it’s where I am in my career, but I publish to share my science with other people. The best way to do that was with a single publication.
BTN: And was that why you decided to request a retraction for the papers that Diao managed to get published?
Sanders: After the Journal of Molecular Biology incident, the journal and university wanted to sanction him and launch a full-scale investigation because you can’t take other people’s data and publish it without their consent. The authorship conflict between me and him was one thing. But trying to publish other people’s data without their consent, that’s something that can result in someone being banned from scientific publication. It’s a serious issue.
But in my foolishness, I said that I didn’t want an investigation and that I didn’t want to punish him. I wanted to publish a paper with him. He made a mistake, he’s sorry, let’s just move forward. And we then had a firm agreement that we would publish one paper.
BTN: He said that he now wanted to publish one high-quality paper?
Sanders: Yes. Absolutely.
BTN: Was this a verbal agreement?
Sanders: It was verbal and in an email. So, I started working on it. I made good progress and sent a copy to a collaborator. Then the phone calls started. It was Diao. He said: “I have to have two papers. I won’t do one paper.” He changed his mind. I said that I wasn’t going to do that.
And I didn’t feel like I could publish the paper without his name on it, so we came to an impasse.
BTN: So, at that point there wasn’t going to be any paper published.
Sanders: That’s right.
BTN: But he went ahead and published anyway.
Sanders: Yes, he stripped out the data from other people and submitted papers. These papers are of very poor quality. The structure is fine, but my strong sense is that he was using my wife’s death to help him get these papers published. So, there’s a combination of a relatively poor article with some of my ideas on the interpretation of the structure. Points that he argued with me and rejected previously were now incorporated into the paper.
When I learned about these papers in the Journal of Bacteriology and Proteins, I contacted the editors to tell them that this was a collaborative project and that he was using data and ideas generated in a collaboration. The Proteins one was held up for print publication, and the Journal of Bacteriology said that Purdue had to do an investigation. Their investigation found that it was indeed a collaboration, that my grants partially funded it, that Purdue owned the data, that I am fully in charge of the data, and that he cannot publish it without my consent. We sent that to the journals, and they retracted the papers.
BTN: So, will there be a paper published on this in the future?
Sanders: No. We are not going to publish. I could get someone to crystallize the protein, redo the whole project to publish it. But that’s not fair to whoever I assign to do that. And it’s not right. It’s a shame that it will not be published, as I said there are a number of interesting elements to the structure, and it’s possible that someone will look at the structure and put it into review papers. If I have the opportunity to write a review paper, I might refer to the structure as it is present in the PDB. There’s just no real good way to go forward.
BTN: Has this changed how you communicate with your postdocs with regards to authorship?
Sanders: No. I have always been explicit with my students on the criteria for publication. I lay it out. I tell them what contributions lead to authorship and what contributions do not lead to authorship. I never had another conflict about authorship with either collaborators or people in my lab. Everyone is aware of my standards.
No, it hasn’t really changed things. You just don’t expect something like that to happen. It doesn’t happen. I mean there are authorship conflicts, but not like this. And if you read the Retraction Watch article on this, he did this to another lab as well. This sort of egregious conduct isn’t all that common. A number of journals do have mechanisms in place to make sure that everybody who is an author agrees to authorship. The American Society for Microbiology is very rigorous. Every author must agree to the responsibility that comes with authorship. That sort of institutional involvement is essential.
BTN: Several commenters on our original news story said that postdocs should have more control over the experiments that they perform. What are your thoughts on that?
Sanders: I don’t disagree that it could be an issue. One of my students who worked on the phosphotransferase was perturbed by those comments. The article didn’t really reflect the complete situation. I wasn’t arbitrarily assigned rights to my wife’s data. So, with that sort of background, it would seem like the postdoc was getting short changed. Here I am, for no reason, objecting to his publication of my wife’s data. It was not just my wife’s data, it was a collaboration. It wasn’t assigned to me after she died, it was mine all along.
But I agree that these conflicts certainly arise. Legally, the data belong to the institution to which the grant has been given. And the institution is most likely to assign responsibility to a full-time long-term employee rather than to a postdoc. That being said, there needs to be a discussion between a postdoc and faculty. It’s an unequal relationship, and sometimes that unequal relationship can be exploited in unfair ways.
At some level, it’s an employer and employee relationship between a lab director and a postdoctoral fellow. Ultimately, it has to be the PI’s decision. He received the grant funding that creates the project. Normally, the postdoctoral fellow did not create the project from scratch. Most of the time, it is much of the intellectual input has proceeded the postdoc joining the lab. And so I would say that the balance has to go with the PI. But it needs to be negotiated. But certainly a postdoctoral fellow—who has conducted research at an institution where he is being paid with the money from grants assigned to the institution and a PI—should not publish the data without the consent of the PI.
BTN: Is there anything that you would have done differently?
Sanders: I generally try to get people to publish before they leave the lab. That’s good practice, but there are a lot of cases where that doesn’t take place. But I try to do that because it’s the best time to do it.
I might have, early on, said this is one paper. This is what it’s going to be. We’re going to do this. In hindsight, I should not have allowed it to go on once he submitted a paper with other people’s names on it. I should have allowed him to be sanctioned at that point. I think a lot of the headache would have been prevented. But that’s just not my style. But when we have somebody who has repeatedly engaged in these things—not only in my lab but in another lab as well—the retractions were important to pursue because the scientific community needs to understand about this individual.
BTN: What advice would you give young scientists who are trying to advance their careers or PIs who might find themselves in a similar situation?
Sanders: Communication is essential at every step of the way. That will reduce the likelihood of these incidents.
But I also have to say that at some point, you have to come to the decision that it’s not going to work out. You shouldn’t continue to try to get something to work out that just can’t. I hate to say it, but it’s like a relationship sometimes. You can try to make it work out, but sometimes all of the clues are already there. It’s simply just not going to work out. You have to decide that there’s no route to a mutually agreeable outcome. And you have to accept that, and move on from there.
Diao declined to comment on the situation in response to a request from BioTechniques.