Important biological discoveries have arrived with the same old-fashioned fanfare for the last three centuries. After months, maybe years of research, a paper will wind its way through the peer review process and land in the pages of (hopefully)…
Important biological discoveries have arrived with the same old-fashioned fanfare for the last three centuries. After months, maybe years of research, a paper will wind its way through the peer review process and land in the pages of (hopefully) a high tier journal: a Nature, a Science, a Cell. Picturing those finalized figures under a glossy cover is enough to set a postdoc’s heart aflutter.
But if it were up to biologists Michael Eisen and Leslie Vosshall, they’d celebrate a paper’s release with a PDF and a rainbow unicorn.
Biologists, publishers, and science funders gathered at a meeting last week in Maryland at the Howard Hughes Medical Institute to talk about how to improve the centuries-old tradition of slow, careful peer-reviewed science publishing. The ASAPbio meeting focused on bringing that process into the Internet age by releasing papers as preprints, skipping over the review process to get research out faster—just as fields like physics already do. And yes, Eisen and Vosshall were promising rainbow unicorn T-shirts to people who took the opportunity to submit their first preprints.
Peer review is scientific policing in action: A journal shares a submitted paper with a group of scientists and returns their comments to the author, who uses those suggestions to make the paper better. That’s the best case scenario. Worst case, that editorial back-and-forth takes months, holding scientific information hostage and slowing the pace of discovery. On top of that, journals tend to publish only the most exciting papers, occasionally preventing scientists from reading about important results that aren’t, well, sexy.
Print journals used to be the only way to share new science. But today, scientists communicate much more rapidly—and directly—than before. “Something’s changed,” says Eisen, who runs a lab at the University of California, Berkeley. “People live on the Internet now. We have senior scientists who are all on Twitter. It just doesn’t make sense that we don’t publish our work immediately.”
Case in point: Back in July, Ron Vale, a biologist at the University of California, San Francisco and one of the ASAPbio organizers, simultaneously sent a paper to the relatively new preprint server bioRxiv.org and the open-access, peer-reviewed journal Proceedings of the National Academy of Sciences. It went up on bioRxiv the next day. It took PNAS a month and a half. In that time, Twitter users sent 600 tweets about the preprint and 6,000 people viewed it. In the same six weeks, Vale got responses from two referees at the journal. “By the time I got this feedback from the journal, it was dwarfed by the amount of feedback that I got through other mechanisms,” says Vale. “I integrated all that information, and it produced a better final article in the journal.”
Still, the biology community—and other fields—have been slow to pick up the practice. Conservative scientists worry that top-tier journals won’t accept their papers for publication if they’ve already appeared on a preprint server. (Nature and Science, for example, will publish papers after preprint, but Cell, the well-regarded biology journal, only does it on a case-by-case basis). Others worry that publishing papers before peer review will let competing researchers scoop their ideas.
What would a new system look like? It’s not “publication.” Is it … release? Posting? Version control, a la GitHub? Nobody’s really nailed down the terminology, or the exact methods. So that’s what last week’s meeting aimed to discuss. “In order for it to be constructive we needed to have anti voices, so it wasn’t just an echo chamber,” says Jessica Polk, a Harvard postdoc and one of the meeting’s organizers.
The 75 invite-only attendees spent two days listening to presentations, including one from Paul Ginsparg, the physicist who built the successful physics preprint server arXiv.org in 1991. Then they broke out into group discussions. “It was exhausting,” says Polk. “The conversations were starting at breakfast and continuing all the way to the closing of cocktail hour.” Some people even submitted their first preprint papers. That’s where the unicorns came in: Eisen gave T-shirts emblazoned with rainbow unicorns to the first five people who posted their first preprint during or after the meeting. (Biologist swag is the best swag).
That rainbow unicorn isn’t just a mascot. Soon, Eisen and Vosshall plan to roll out a server that will integrate preprint studies with post-publication peer review—kind of like the feedback that Vale got on his bioRxiv paper. It’s the same model that a number of new publishers, including PeerJ and F1000 Research, are trying to promote. Science is supposed to be an iterative process, with self-corrections every stage of the way. The rainbow unicorn—obviously—stands for making that reality match the way that science gets disseminated.
It could also stand for a larger shift. Today, when you read about a study published in Nature, you’re likely to take it at face value. It’s in the most prestigious science journal in the world. Editors have vetted it and scientists have reviewed it. It’s got to be true, right? Wrong. “A paper really isn’t the scientific community saying, ‘we’ve decided this is correct.’ It’s a scientist putting forward their work.” says Eisen. Saddle up that rainbow unicorn, though, and that work really sings.
Click here to view full article