In two papers Oste and her group were analysing,guided by research integrity detective Professor Jennifer Byrne and PhD student Pranujan Pathmendra,they stumbled across five cell line names they couldn’t verify. Cross-checking cell lines used in experiments should be easy using a comprehensive online encyclopedia called Cellosaurus.
“When we found these five[non-verifiable cell lines],I just remember the Cellosaurus page coming up saying,‘no results’,and thinking like,what? What do you mean? No results,how?”
Danielle Oste,who’s finishing off her animal science degree at the University of Sydney,travelled to Athens to present at a conference on research integrity.Credit:Danielle Oste
Byrne called Oste back to continue research on the issue,outside of class. They found more phantom cell lines – 23 in total – and zeroed in on seven that have been referred to across hundreds of papers.
These cell lines weren’t in Cellosaurus. There were no previous papers that described how the cells were established. They were missing genetic identification markers. And,although some researchers claimed they had obtained the cells from repositories (online stores for lab ingredients),Byrne and Oste couldn’t find them listed in catalogues.
Some of the missing cell lines may be the result of human error,such as researchers misspelling cell line names. But misspelling alone cannot explain the phenomenon in more than half of the 420 papers they analysed,the researchers say.
“I know it sounds terrible,but I don’t think these cell lines exist,” Byrne says. “We think we’ve discovered the tip of a much larger problem in the scientific literature.”
Why does it matter?
The existence of these phantom cell lines is chilling because it undermines critical medical research. Scientists need to be able to reproduce and verify each other’s work or look at results from cells and feel confident trying to reproduce the results in real organisms or humans.
“The concern that I have is that many laboratory researchers are actually out there possibly trying to reproduce experiments that never existed,” Byrne says.
“How do you reproduce an experimental result in a cell line if you cannot get hold of that cell line? The answer is:you can’t.”
The missing cell lines have also contaminated literature reviews,the gold standard of scientific evidence that collate hundreds of studies to draw robust and reliable conclusions. As unwitting scientists refer to studies with unverifiable cell lines,the cell lines get ironed into the literature and infect our understanding of human biology itself.
Professor Jennifer Byrne,a cancer expert and research integrity sleuth at the University of Sydney.Credit:Wolter Peeters
“The reach of cell lines forward into the literature is very long because they’re these foundational models that everyone uses at the start,” Byrne says. “If the foundation is rotten,then what you can build on that is not going to hold.”
What now?
Calling out the specific authors who used these non-verifiable cell lines was outside the scope of Byrne and Oste’s research,but they’ve written to affected journals to flag their concerns.
Loading
“Whether they start off as misspellings and then seem to take on new lives as independent cell lines,or whether other stuff’s happening,potentially,with AI or paper mill involvement,it’s hard to say,” Oste says. “But there’s some definite red flags there.”
Many of the problematic studies emerged from Chinese hospitals,which have been identified as a hotbed of “paper mill” activity. This refers to researchers churning out papers with shoddy data in disreputable journals just to boost their own professional standing or attract funding.
“We’re not trying to point fingers,but we just want to make it clear that this is of significant concern,” Oste says.
“Papers that are describing experiments with these non-verifiable cell lines should be retracted,and any literature reviews or other papers that refer to them should have expressions of concern.”
Oste,still an undergraduate,has just returned from Greece where she presented at a research integrity conference,ahead of her exams. Byrne says their paper reporting the missing cell lines,in theInternational Journal of Cancer,is a shining example of student-led research.
“Students who just have no preconceptions about what’s out there can actually make very powerful observations,” Byrne says. “Possibly hundreds of thousands of people have seen these identifiers,and no one has thought to check them – until we did.”
Examine,a free weekly newsletter covering science with a sceptical,evidence-based eye,is sent every Tuesday. You’re reading an excerpt –sign up to get the whole newsletter in your inbox.