Preprint papers posted on servers such as arXiv and bioRxiv are designed to get research results out for discussion before they are formally peer reviewed and published in journals. But for some scientists, the term is now a misnomer — their preprint papers will never be submitted for formal publication.

Graham Coop, an evolutionary geneticist at the University of California, Davis, took to social media on 13 January to state that one of his bioRxiv preprints is the “final version”.

Coop’s paper1 critiques a 2015 study about a process known as genetic hitchhiking — changes in the population frequencies of gene variants owing to their close proximity to other gene variants that are increasing in frequency as a result of selection.

Part of his reason for leaving the paper as a preprint is because it is a response to a previously published article, rather than a full article in its own right, says Coop. But another reason for the move, he explains, was to experiment with how preprints are perceived by researchers.

David Stern, an evolutionary biologist at the Howard Hughes Medical Institute's Janelia Research Campus in Ashburn, Virginia, noted that his group already follows this ‘pre-as-final’ path for some papers.

One of the major services of traditional journals is that papers are peer reviewed before publication, allowing authors to make changes in response to referees' comments. Increasingly, some preprint advocates suggest that readers informally peer reviewing papers after they are posted — a form of post-publication peer review — can substitute for this.

Vincent Lynch, a human-genetics researcher at the University of Chicago in Illinois, supported Coop, and said that peer review after publication was more important than traditional journal review beforehand.

Lynch thinks that the idea of 'final version preprints' will catch on with other researchers. “Publication in a journal doesn’t magically transform data from conjecture into fact,” Lynch told Nature. He notes that readers of journals still judge the quality of the data in papers, which can be done just as easily with final version preprints.

But, perhaps surprisingly, Coop tweeted that he doesn’t see the final version preprint as likely to catch on widely just yet — at this point, there isn't enough post-publication peer review (or “PPR”) so he still sees a use for peer review (“PR”).

Lynch and Coop also caution that using preprints as final versions may not work for younger colleagues, who need to show a track record of journal publications to help advance their careers. “Students and postdocs will have likely have job and grant applications reviewed by those who don’t share my view or take preprints seriously,” Lynch adds. “It would be unfair to them to only submit a preprint.”

Coop was the only author of his paper, and he says he would have probably submitted the preprint to a journal if his students contributed significantly to the work.

Independent data scientist Jordan Anaya, who has developed a search engine for preprints called PrePubMed, tweeted that as more final version preprints appear, there will need to be an easy way to show online which ones are actually final papers.

Anaya told Nature that he hasn’t submitted his last two preprints to journals, noting: “If the work is important enough that people read it, hopefully someone will point out if they disagree with it or find something wrong with it.”