IIn 2020, First-year graduate student Laura Lubert was asked to present two classic papers to her journal club about the honeybee waggle dance, which individual bees use to communicate with other bees in the hive about the direction and distance of food. Lubert was new to the subject, so she decided to read up on additional research. Recent Blog PostAs she read, she wrote, “something felt strange; I felt like I was looking at the same data over and over again.”
Lubert ended up teaming up with his Caltech supervisor, computational biologist Lior Pachter, and the two analyzed a series of papers about honeybees co-authored by the well-known scientist Mandhyam Veerambuddy Srinivasan. They ultimately found what they characterized as “problematic behavior across a number of papers.”
When graduate student Laura Lubert discovered the problematic research on honeybee waggle dances, at first, “nobody wanted to talk to me,” Lubert says.
Visual: Courtesy of Laura Lubart
Lubard and Pachter published their findings in a paper. Preprint The server was shut down in May. But their concerns Draw a lot Note By July, Pachter was now widely viewed. thread On Platform X, formerly known as Twitter, he called honeybee research “garbage.”
Srinivasan later responded to the pair’s blog post, criticizing them for conducting their investigation in an “unprofessional manner.” Instead of reaching out directly, he wrote, “they drew conclusions based on a limited understanding of the scientific field they were investigating.” Such behavior, he wrote, “does not help advance science.”
Lubert, who recently completed his doctorate and will begin work as a postdoctoral researcher at the Broad Institute this fall, spoke about the controversy and his years of work to correct what he sees as flaws in the scientific literature in a Zoom interview. The conversation has been edited for length and clarity.
Andark: As a graduate student, it must have been unsettling to read papers by world-renowned researchers and question the data. Do you remember your thoughts and feelings at the time?
Laura Lubert: “I was really shocked to see that we were talking about the exact same data. So I spent a lot of time trying to make sure that I was interpreting the experiment correctly. I wanted to discuss my findings with other scientists, especially those who are experts in insect behavior.” [present to] My journal club had a very detailed discussion section where I showed all the instances of duplicate data that I found at the time.
What was really shocking to me, and what really stuck with me and became a frustration for the rest of my PhD, was the reaction I got. No one wanted to talk to me about it. There was no interest or surprise. The senior scientists that I contacted were like, “Oh, that’s how science works.”
UD: The scientists you spoke to directly didn’t seem interested in this subject. What have you done to spread the word?
LL: I tweeted about it. I barely had any followers at the time, so it didn’t get very far. But the comment that’s really worth mentioning was by Dr. Elizabeth Bik, who really Career Her job is to flag problematic papers. She actually helped me find these papers on PubPeer and encouraged me to write comments there. So I wrote anonymous comments on PubPeer at the time, and they were completely ignored by the authors and the journals, but at least I thought, “Okay, the paper is out there.”
UD: What is PubPeer?
LL: PubPeer Website You can add comments to individual papers.
UD: Eventually you joined Lior Pachter’s lab, and the two of you decided to really look into this line of research. Why do you think no one had done this before?
LL: It’s not that easy. It’s very unappreciated work. We spent a lot of time on these papers, reading them, trying to understand them, trying to write very objective reports about what we found and what we thought were the problems. This kind of work is just time-consuming and not worth it for most scientists. It’s not appreciated at all. On the contrary, I was told that speaking out about these problems could have a really devastating effect on my career.
UD: How much time did you spend on this endeavor?
LL: A lot. When I was setting up the first journal club, the pandemic was just starting, so I was stuck at home anyway. I really worked on it full time for a month.
Actually, this has nothing to do with the question you asked, but the reason we revisited the thesis is that this frustration stayed with me so much that when Lior and I were talking about what I wanted to do after my PhD, I mentioned this as a reason why I thought academic science might not be for me.
Lior’s reaction was completely different from any reaction I’d had before. He immediately said, “Wait a minute. You found this and nobody cared? It wasn’t published?” I showed him the findings and he was really shocked. Then, as we read through the paper together, we actually found a lot more problems.
Since October, Lior and I have been working on the manuscript.
UD: Wow, it’s July now. I submitted the manuscript to —
LL: In May.
UD: That’s a long story. Did this work compete with your doctoral studies?
LL: Yeah, definitely. I was trying to get a PhD, but it has nothing to do with insect behavior whatsoever.
UD: In your blog post, you talk about your efforts to contact publishers of your honeybee paper and to post your manuscript on a preprint server. It doesn’t seem like an easy process. Why do you think journals and preprint servers have been reluctant to acknowledge your findings?
LL: First, let me explain why journals are reluctant. Unfortunately, the incentives for journals are currently misaligned. Journals have their own peer review process, but once a paper is published, journals don’t want to admit that there might be something wrong with the paper they published. This goes against how science works, because we question things and we’re sometimes wrong.
Obviously, in this case we are talking about much more egregious errors, but journals are very reluctant to admit that there was anything wrong with the papers they published unless the problem was flagged up in their own peer review process.
For preprint servers: BioRxivThey have a very clear rule that they only publish new research, which I understand. That’s fine, but I think it’s clear that that means that this kind of research needs a different platform. And what Lior and I realized after writing this manuscript is that it’s really hard to publish this kind of research. There really isn’t a platform.
We are truly grateful arXiv I am happy to be able to publish the manuscript there. But I also wrote a blog post about it because I wanted to be able to leave comments on the manuscript. All this shows that we need a better platform to publish this kind of work.
UD: If a researcher feels there is a problem, is the first step to contact the study authors, the journal, or anyone else?
LL: I don’t think that’s a solved problem. That was the problem I had. Even senior professors at Caltech didn’t know where to send me.
UD: Mr Srinivasan said that you and Mr Pachter conducted the investigation in an “unprofessional manner”. How do you respond to that?
LL: I don’t know exactly what he considered unprofessional, so it’s a bit difficult to respond to that. If you read our manuscript, you will see that we took great care not to name anyone and to communicate our findings in the paper very objectively. The reason we wrote the blog post is to provide a platform for scientists and experts, including Mr. Srinivasan, to react and discuss the issues we found. We also made sure that all of our analyses were very transparent and reproducible.
UD: Do you think peer review should or could be the forum for catching these discrepancies? Do most reviewers look at a researcher’s previous work to see if the data or methods have changed from study to study?
LL: What Lior and I did was peer review; it was just done outside the peer review process confined to the journal. I don’t think it’s reasonable to expect journal reviewers to look at the entire study every time they review a paper, because they’re doing the peer review. unpaid.
But I think we need a platform for post-publication peer review, because when these problems are discovered pre-publication, the way journals respond is very different than it is post-publication, where journals, for some reason, tend to try to sweep all the problems under the rug.
UD: What would you like to see happen in the future?
LL: I’d like to hear a more objective answer. [honeybee research] The authors will see if any of the things we’ve discussed can actually be explained, or if any of the data are still available and the analysis can actually be redone.
UD: Do you and Pachter plan to submit new manuscripts to journals?
LL: We haven’t talked about that yet. Our main goal was to correct the scientific record. It’s very unfortunate that the tweet took so long to get there, but at least now there’s been enough public response that the journal was forced to investigate it.
And, more importantly, our findings have been shared broadly within the entomology and behavioral communities, but it would be nice if this paper could be peer-reviewed and cited as published in a peer-reviewed journal.
UD: In your blog post, you both suggested a journal dedicated to reviewing bodies of research by scientists or teams of scientists. Can you tell us a bit about that?
LL: The Journal of Scientific Integrity was born out of the idea that there needed to be a place to disseminate this kind of research. We wrote the manuscript and realized there was no place to disseminate this research at all. Interestingly, [X]In fact, we received a lot of responses from scientists about how they would have tried to fund such a journal, but unfortunately there was very little interest.
UD: Is there anything else you’d like to add?
LL: Careful review of research findings is crucial to scientific progress. In many cases, it saves lives. We have seen cases like this: TheranosDr. Elizabeth Bik Unveiling Ceremony lots of Problematic Jobs Professor Didier Raoult is a researcher who actually sees patients, so he literally puts his life at risk when reviewing problematic studies.
Of course, in the case of Srinivasan, he is not working on human health. But I still think that careful review of papers is very important for science. And that applies to our paper too. So anyone, including Srinivasan, can and should carefully review and comment on our paper, especially if we find that we did something wrong. I think the essence of science is to try to find the truth together.
