Jennifer Byrne leads global hunt for fake research papers

Jennifer Byrne has written an editorial on the topic, published last week in the prestigious journal Nature.

Jennifer Byrne, professor of molecular oncology at Sydney University. Picture: Simon Bullard

An ingenious computer program, and generous grant funding, have helped Jennifer Byrne track down potentially fraudulent research papers. So far she has found about 160 of them.

A professor of molecular ­oncology at the University of Sydney and head of the Children’s Cancer Research Unit at the Children’s Hospital at Westmead, Byrne’s career has been focused on analysing childhood and adult cancers at a molecular level, and she has published a quantity of research on genes connected to cancer. Along the way, she became interested in fake research papers.

It all began about 20 years ago when Professor Byrne published a paper on a specific gene, and then put the work on that gene to one side to work on other projects. Fifteen years later, in about 2015, she noticed five oddly similar ­papers had been published on that same gene.

From researchers at different Chinese institutions, the papers ­included basic (and suspicious) mistakes in the nucleotide sequences they quoted. For instance, from 2000 bases in the gene, some of the authors chose to concentrate on exactly the same 21 bases.

“At the beginning we didn’t have the faintest idea what was going on,” Professor Byrne says. “I had never thought about any of these kinds of things in my life.”

But on further investigation, she determined that experiments as described in the papers simply could not get the results specified.

“If you had done the experiments in the way they had described, you cannot get those results,” she says.

The revelations continued. “Now we’ve found other, similar papers,” she says. “At the moment, we have about 160 papers that probably have sequence errors. Once we started looking, we started finding papers that had these kinds of features, maybe as far back as 2010.”

She wrote to the authors of 48 of the suspect papers. None ­of them replied. She also contacted the journals, resulting in 17 retractions to date.

Now she has written an editorial on research fraud, published last week in the prestigious journal Nature, and co-authored a paper on the subject for the journal ­Biomarker Insights, which discusses fraudulent research papers and the damage they can cause to ­potentially lifesaving scientific ­research.

Professor Byrne and her colleagues are screening papers from nations other than China as well, and they have found some with ­sequence mistakes, but of a more ordinary, human-error type: the inadvertent transposition of ­sequences between two genes, or the jumbling of the letters of some sequences.

Others, though, are more extreme and more difficult to explain. “There are many of these kinds of papers in the literature and at the moment we are scratching the surface,” she says.

Professor Byrne used the $US150,000 ($211,000) grant funding she was awarded last year from the US Office of ­Research Integrity to hire a ­research assistant to forge ahead with tracking down potentially fake papers.

Professor Byrne’s colleague in France, computer scientist Cyril Labbe of Grenoble Alpes University, has developed a tool to assist them in the hunt. This computer program tool, Professor Byrne says, “can crawl across the text of a paper”, tracking down errors and inconsistencies.

“It knows what a nucleotide ­sequence looks like, because a ­nucleotide sequence only has a limited number of characters in it,” she adds. “It can pick the words that only have those letters in them, with length limitations. It can look around for a description of the ­sequence and pull that out as well. It then identifies the sequence via a database according to all the ­research done on genes until now, and it looks for sequences that have been incorrectly classified.”

The tool is freely available for researchers who would like to use it in their own investigations. She and Dr Labbe have written a paper on the tool and its applications, which is now under consideration by a journal.

Professor Byrne believes the scale of the fraud being revealed could be the result of universities demanding scientists publish frequently. “The scope of research fraud that we’re uncovering is exceeding people’s expectations,” she says. “It’s frightening, it’s really frightening. It’s creating a (skewed) evidence base that ­people will then use for further, more clinically directed research.”