When three of the world’s biggest private biomedical funders launched the journal eLife in 2012, they wanted to shake up the way in which scientists published their top papers. The new journal would be unashamedly elitist, competing with biology’s traditional ‘big three’, Nature, Science and Cell, to publish the best work. But unlike these, eLife would use working scientists as editors, and it would be open access. And with backers providing £18 million (US$26 million) over five years, authors wouldn’t need to pay anything to publish there.

Four years and more than 1,800 publications later, eLife’s funders — the Howard Hughes Medical Institute in Chevy Chase, Maryland, the Wellcome Trust in London and the Max Planck Society in Berlin — announced on 1 June that they will continue their support. They will back the non-profit eLife organization with a further £25 million between 2017 and 2022 (see ‘eLife by the numbers’).

eLife’s status in the field is rising quite quickly,” says Sjors Scheres, a structural biologist at the Laboratory of Molecular Biology in Cambridge, UK. He became an editor at the journal in 2014, overseeing papers on electron microscopy. “I liked the idea behind it — to make a high-impact journal completely driven by scientists, and open,” he says.

But although scientists like publishing in the journal, it’s less clear whether it will drive wider transformation at the elite end of science publishing.

boxed-text

Collaborative attraction

eLife has gained high-profile converts. Palaeoanthropologist Lee Berger at the University of Witwatersrand in Johannesburg, South Africa, is used to publishing descriptions of the fossils he finds in Science and Nature. But his team’s latest find — more than 1,500 remains attributed to a newly designated species Homo naledi — became eLife’s first palaeontology papers in September, and Berger hopes that more are on the way. “Our next big discoveries are going to appear in eLife — if they’re accepted by the referees,” he says. “If not, we may submit them to Nature or Science.”

Huda Zoghbi, a neuroscientist at Baylor College of Medicine in Houston Texas and an editor at eLife, says that her postdocs — soon to be on the job market — still feel the need to publish in the big three journals. But she thinks that attitudes are changing. With some work, “People ask me ‘Why did you send it to eLife? You could have sent this to one of the top three journals’,” says Zoghbi. “I told them eLife is a top journal.”

The journal’s most innovative feature, according to its authors and reviewers, is its collaborative peer-review process. It turns conventional peer review — in which referees submit individual, and sometimes contradictory, reports — on its head. Instead, referees and scientist–editors work together to identify a submitted paper’s strengths and weaknesses and any needed revisions. Authors receive one decision letter, not individual reports from each referee.

That makes for a speedy review: last year, eLife’s published papers took, on average, 116 days from submission to acceptance. For comparison, Nature and Cell take around 150 days, although Science says that in 2013 it took 99 days from submission to acceptance. Cell and two of its sister journals have experimented with a similar peer-review model but none has yet adopted it. Pete Binfield, the publisher of another open-access journal, PeerJ, in San Francisco, California, says that he likes eLife’s peer-review system, but he thinks that the approach would be impossible to scale up to adopt for all published articles.

Selective but open

As it bids to become a top journal, eLife has started to turn down more of its submissions. The journal’s acceptance rate dropped from 26% in 2014 to 15.4% by 2015, says its editor-in-chief Randy Schekman, a cell biologist at the University of California, Berkeley. That’s approaching the acceptance rates of Nature and Science, which are both below 10%.

In 2013, Schekman denounced Nature, Science and Cell as “luxury journals”, and likened their low acceptance rates and high impact factors to high-end “fashion designers” that artificially stoke demand for their brand through scarcity. Now, he says, eLife has become “more selective than I had imagined, but it’s not based on any instructions I have conveyed to the editors. It’s based on their sensibility of important work.”

eLife’s funders hoped to show that a highly selective, innovative journal could also be open access. When in 2014 the selective journal Nature Communications, which is owned by Nature’s publisher, made all its articles open access, a spokesperson for the publisher cited eLife’s existence as an influence.

But eLife has also demonstrated that such a model costs a lot of money. In 2014, the most recent year for which financial information is publicly available, eLife published 537 research articles with expenses of £3.4 million — equating to around £6,300 for each article. “It appears to be a very expensive way to innovate in the publishing space,” says Binfield.

The journal says that its per-article cost has dropped — to £3,522 in 2015. It points out that it spends money on technology development, too; it has launched a series of open-source publishing tools to manage peer review and to display articles. Six publishers that use the third-party publishing platform HighWire have tested the eLife-developed Lens display technology, but Binfield says he doesn’t see people flocking to use eLife’s platforms.

Schekman says that eLife plans to diversify its income by asking governments and other charities for funding. It will also eventually charge scientists to publish in the journal, although Schekman doubts that such income will ever cover all the organization’s expenses.

But it won’t, he says, establish other open-access journals that accept more papers and have lower selectivity — a strategy that some organizations, such as the Public Library of Science, or PLoS, have demonstrated can shore up finances. “We have no interest in creating other lesser journals with lower standards,” he says.