On a Tuesday afternoon in the spring of 2013, Richard Sever logged on to Twitter and wrote a brief post introducing the world to bioRxiv, “biology’s answer to the arXiv preprint server,” as he put it. 

It was a fitting place to let life scientists know that the Cold Spring Harbor Laboratory Press, where Sever was an assistant director, was launching an online repository of preprints modeled on the popular arXiv, where physicists routinely posted preliminary drafts of their research papers. Twitter was where Sever had first seen discussions among genomics researchers about whether the field should embrace the free sharing of manuscripts prior to their formal publication in peer-reviewed journals. 

This practice was still slow at that time to win support among biomedical scientists. Some feared being scooped on discoveries with potential applications in drug development. Others worried that unvetted research could influence public health practices. One of the biggest concerns was whether posting a preprint could jeopardize future publication in one of the top journals, which in biomedical fields wield undue influence over researchers’ career prospects. 

More than a decade later, some of those fears still persist, but the uptake of preprints among life scientists has gained serious momentum. About 14% of published biomedical research now exists first as preprints, according to data from Europe PMC. Roughly half of those appear on bioRxiv or medRxiv — a medical and health science-focused server launched in 2019 — Sever said. Together, they draw 10 million visitors each month. 

Use of the servers has grown so much, in fact, that this month, Sever will be leaving Cold Spring Harbor to become the chief scientific and strategy officer at a new nonprofit whose sole mission will be ensuring the future of bioRxiv and medRxiv. The organization, called openRxiv, will be supported by “substantial funding” from the Chan Zuckerberg Initiative, and financial commitments from the Sergey Brin Family Foundation and a number of scientific institutions in the U.S., U.K., and Europe. 

In an interview with STAT, Sever talked about the evolution of preprinting in the biomedical sciences over the last 11 years and how they’ve changed the pace and practice of research. His comments have been edited for length and clarity.

What’s behind the launch of this new venture, openRxiv? Why now?

Both bioRxiv and medRxiv, in many respects, were an experiment. When we launched, I was hopeful, but I wouldn’t have bet my house on it being a success. But it has been — that was particularly evident during the pandemic. Here we are many years later with one-third of a million preprints. Cold Spring Harbor has been a fantastic steward, but it’s matured beyond something that’s simply part of a parent academic institution. It’s become this global thing that is of undoubtful value to the scientific community, and that should be reflected in having it be its own separate organization. 

What will the change mean, practically, to users?

This is not a commercial spinoff, it’s an independent nonprofit and the mission is going to be the same. So for most users there’s not going to be any discernible change in how it’s used. It will still be free. We’re building out a broader governance structure with a board of trustees and a scientific medical advisory board and we’ll be appointing a CEO — someone experienced in nonprofit leadership. Part of the new role of the leadership is to develop a long-term revenue plan. We want to diversify that to ensure long-term sustainability so that we can keep doing the same things we’re doing now as a free service. 

When you and I first spoke, not long after bioRxiv’s launch, it seemed like biology was at a tipping point — preprint servers were starting to catch on in some corners but it hadn’t yet become part of the scientific culture. Where do you think we are now?

I think the benefits of rapid dissemination have become inarguable. Scientists have been voting with their feet. We saw it first in physics, in math, in computational science, now the biological sciences. In every field more people are posting preprints. Clinicians are still a few years behind — there’s still some cultural change that will need to happen there to move from the early adopters to the mainstream. But a lot of those “are there perils?” conversations have kind of gone away. 

Before, a lot of the rationale and the logic behind the mission was in some ways hypothetical; if everybody posts work early, then good things might happen. That has all come true. Every time I go somewhere and give a talk about bioRxiv, a young, early-career principal investigator comes up to me and says that they have their job because of a paper they were able to put on bioRxiv — that if it didn’t exist they’d still be a postdoc waiting on a paper to be published in Cell. So to me, that demonstrates there’s been this really tangible effect on individual scientists and their careers. 

Are you seeing that buy-in from institutions and funders as well?

Yes, definitely. I’ve spoken to three Howard Hughes Medical Institute investigators who said they were renewed because of papers they’d put on bioRxiv. Rockefeller University has encouraged people to put preprints on their job applications. So it’s becoming a norm of the ecosystem now. And there’s this recognition that you can judge the quality of people’s work before it gets the label of some journal and where it is in the pecking order. That’s a real change.

I’ve also seen research institutions using preprints to train Ph.D. students and postdocs on how to do peer-review for the first time because now they have a substrate that hasn’t had all the kinks worked out. That’s been really encouraging to us because we’d always hoped that bioRxiv and medRxiv would stimulate evolution in the community. 

You mentioned the Covid-19 pandemic as another place where the value of sharing knowledge quickly and allowing researchers to rapidly build on each other’s work became apparent. I remember trying, and failing, to keep up with the deluge of preprints that were arriving every day during those years. What was it like for you, being on the other side of it?

It was a very crazy time for all of us. You know, we don’t peer review papers or make editorial changes. Our policy for content screening is to look at three things: scope, relevance, and danger to the public. But there were so many papers, my colleagues and I were working 14 hours a day, seven days a week just to process them all. What was slightly surreal would be when I would see a paper on Sunday night and then on Monday there would be an article in the New York Times about it. What was more surreal was that it kept happening. 

In the height of the pandemic, we started having a call with all the founders of medRxiv every Tuesday morning at 9 a.m. because there were so many papers we had to make decisions about. We did that every week for like three years. And I remember that we had this worry, about what happens if one of us gets sick? How does it keep moving forward? 

I think there are some echoes of that feeling now as we make this transition. Seeing how the speed of preprints are impacting people’s careers, it feels very important to make sure we can keep doing this for the long haul. 

You’re making this transition during an interesting moment in biomedical science — as the Trump administration is taking drastic steps to reshape how such research is both funded and communicated. In this new political climate, critics have pointed out that preprints are adding to the public relations challenges for vaccine safety research, as an online furor over a paper on medRxiv last month about Covid vaccines demonstrated. I’m wondering how you’re thinking about the role of preprint servers at this moment and what sort of responsibility you feel to safeguard against the potential that papers that appear on them get weaponized?

In terms of preprints and public health, we continue to pursue our responsible screening approach for papers that we feel might introduce a public health risk, and we have turned away a number of papers for these reasons. 

In the case of the recent medRxiv paper, our judgment was that this did not introduce a significant risk as the phenomenon was already known and had already been reported — the paper merely tried to provide further details of a known phenomenon, rather than discovering it or claiming a higher incidence. In fact, it made many efforts to stress the importance of vaccination. 

It is unfortunate that things are being further weaponized in the current climate, but this applies to both journal articles and preprints, and these actors will leap on anything and misrepresent it, regardless of what is actually said, and it would be dangerous to avoid disseminating literally anything that included vaccine data.