Open Access for Open Science

The EMBO Journal and EMBO Reports join EMBO Molecular Medicine, Molecular Systems Biology and Life Science Alliance as Open Access journals from 2024. Full Open Access at EMBO Press completes another step towards the goal of an integrated Open Science approach for the dissemination of highly selected and curated science.

The EMBO Journal and EMBO Reports join EMBO Molecular Medicine, Molecular Systems Biology and Life Science Alliance as Open Access journals from 2024. Full Open Access at EMBO Press completes another step towards the goal of an integrated Open Science approach for the dissemination of highly selected and curated science. EMBO Reports (2023) 24: e57638 T he value of open data sharing with minimal delay was exemplified most prominently by the publically funded human genome project at the turn of the century, which both facilitated J. Craig Venter's shotgun sequencing assembly published coincidentally, and also ultimately benefited from the Celera project. As Venter noted later, "We absolutely used the mapping data that the entire human genetics community accumulated. . .had there not been a genome project . . . there never would have been a Celera sequence".
Contemporaneously with the human genome milestone, openness reached the publishing world with the foundation of PLOS in 2000, which spawned its first journal PLoS Biology in 2003(Pariente, 2023 as well as with the definition of Open Access (OA) by the Budapest and Berlin declarations in 2002 and 2003, respectively. EMBO's Molecular Systems Biology was conceived in 2005 as an OA journal, reflecting the ab initio open principles of the systems biology community.
The EMBO Journal and EMBO Reports have provided an OA option to authors since 2007. The broad scope and global reach of both journals means that authors have diverse levels of support for OA publication and can thus choose their mode of publication accordingly. A key principle at EMBO Press is-and will remain-equity. We interpret equity as providing realistic access both to readers and authors without compromising on the value that the EMBO Press journals add to the scientific process through peer review, editorial selection, editing, quality control, and curation. The costs of quality publishing can appear high -prohibitively so for some authors. In an effort to add transparency to the costs of the publishing process, EMBO Press has published its finances since 2019, which does not in fact include investment into open science Projects such as Review Commons and Source Data. The costs derive partially from our involved selection and quality control processes. For example, at EMBO Press editors spend about 17 h on a paper that ends up being published and just under 2 h on most papers that are not peerreviewed (https://www.embo.org/features/ the-publishing-costs-at-embo/).
Elective OA (often referred to as "Hybrid Journals") is thus a pragmatic mechanism for the community to choose the mode of publication based on their own priorities, support, and mandates. Beyond ensuring access to all authors, the elective OA model also accommodates three other challenges with the "author pays" OA journal model through subscription income: i In highly selective OA journals, the authors of published papers effectively subsidize the selection process of an order of magnitude more submissions (EMBO Press journals publish between 10 and 15% of submitted research manuscripts). ii The authors of scholarly article types other than research papers and reviews rarely benefit from the support of OA publishing funds and consequently OA journals cannot expect to charge for such articles and instead have to rely on cross-subsidy from research papers.
iii "Author pays" OA has resulted in strong downward pressures on price, providing a challenging environment to invest in quality assurance, curation, and open science processes at progressive journals. For example, EMBO Press currently invests over 300 k € annually in quality control beyond the editorial/ peer review processes.
The challenges of ensuring that Open Access improves scientific dissemination are evidenced by the massive proliferation of "low threshold" OA publications, often with limited added value beyond a brand and an online platform-current estimates suggests there are up to 40,000 journals in the science, medicine, and technology sectors. This does not take into account the dramatic rise in the so-called papermills, illicit platforms that extort a system over-reliant on misleading metrics related to journal publication for research assessment. The explosion of "no frills" OA journals in the last few years should concern anyone interested in an effective scientific process that efficiently invests public funds in research progress. As a community, we have the choice to support quality publishing by submitting judiciously, and we should not lose sight of investing where value is added-with an eye on promising Open Science developments.
From OA papers to OA journal EMBO Molecular Medicine was "flipped" from subscription to full OA in 2012-the first highly selective OA biomedical journal to do so. It preceded the conversion to full OA of the remaining two EMBO scientific publications by a dozen years for two reasons: First, the global scientific community and the public has a moral right to access peer-reviewed research with direct biomedical relevance without restriction; second, the size differential of the journals meant that EMBO could subsidize its smaller journals with subscription income from its larger journals to ensure equitable OA charges. Nonetheless, we always viewed elective OA as a mechanism that would ultimately selfadjust to full OA once the majority of authors were ready to choose this mechanism of publication. It appears that this time has come: Fig 1 shows that the proportion of OA research papers at both journals has grown steadily since 2017 to reach over 60%. While these data look straightforward at face value, it hides dramatic differences in research funding and OA policies across the globe. The Open Access choice at the EMBO Journal and EMBO Reports is strongly enriched with European authors, reflecting the prevalent OA charge support infrastructure and emerging OA publication mandates by many EU-based funders.

Show your colors
The good news is that in the biosciences the move to global OA is now well underway also beyond EMBO Press. The bad news is that the financial support for this conversion remains as heterogeneous as the current funder OA mandates. Europe has largely embraced "gold OA"-that is full OA publication with a CC-BY license and author copyright retention. This is supported by a number of key policies including by the Council of the European Union and the European Research Council (ERC), as well as widespread "Publish&Read" or "Transitional" agreements, such as the nationwide DEAL consortium in Germany. Nonetheless, transitional agreements often exclude full OA journals and DEAL is currently set to expire later this year. Together with a lack of synchronization between national OA funding schemes in Europe, this conspires to general lack of opportunity for secure financial planning for OA publishing. Coalition S, a consortium of 26 major life science funders, brought some international coordination and, with a start date of 2024, added urgency to many journal conversion plans, but notably restricts itself to present a general policy framework without financial coordination.
Moreover, North America, epitomized by new policies by the White House Office of Science and Technology Policy, the US National Institutes of Health (NIH), and National Science Foundation (NSF) focuses on Open Science mandates with collateral support for "green OA" (that is posting of a freely accessible preliminary version of a research paper on a repository). Finally, many Asian and South American countries have signaled that they in principle favor Open Science and/or Open Access but are yet to provide consistent national policies and dedicated financial support.
All EMBO Press journals are fully compliant with the Open Science and Open Access policies listed above. EMBO unequivocally supports the "gold OA" route as it best preserves the important principle of a reliable scholarly record. However, the overarching goal has to be a much broader Open Science landscape for the dissemination of reliable scientific results. Open Access cannot be seen as an end point. This is echoed by the newly released EU council recommendations: 'OA is one of the core elements of Open Science'. Indeed, we believe that quality journals can and should play an important part in Open Science developments as proven providers of peer review, quality control, research integrity and curation services, as well as enrichment, discoverability, and contextualization-crucial in an overwhelmingly complex research landscape. Advocates of dissemination platforms without these attributes fail to appreciate the risk of drowning in a mass of uncrated, uncontextualized information. They also fail to understand the importance of quality journals in encouraging transdiscipline exchange (although we still struggle to emulate the eminent browsability of the cellulose world in the digital medium).
Thus, OA is but a component of the larger Open Science goal. The latter in fact requires coordination and broad additional investment by funders, research institutions, and indeed journals. Those who see a primary function of OA to cut the charges levied by commercial publishers miss the point that this will also undercut critical Open Science infrastructure investment. It would be wise to analyze the economics of these developments more formally, starting with estimates on the value proposition of "green OA"-in my view, the uncoordinated proliferation of unstructured institutional repositories in a scramble to meet government policy mandates is in general not an efficient investment of research resources-and moving on to estimate the funds diverted to journals that add minimal value beyond online publication.

Fair OA
There are numerous challenges to a financially equitable OA model that relies substantially on author payments, which remains the most prevalent and scalable approach at this time. The most fundamental is that concentrating costs on two to three orders of magnitude fewer authors than readers will amplify inequities in the global research community. These include differences in research funding and publication support across disciplines and nations, sometimes exacerbated by unfunded OA publication mandates. Furthermore, only a subset of-at this time mostly European-countries have shifted the existing journal subscription budgets at academic libraries successfully to OA publication charge resources. This is a crucial but complex task, as the economic burden substantially shifts from "read institution" (subscriptions) to "research institution" (OA). For the rest of the world, typically authors have to cover charges from research budgets, which are not designed to fund publishing. The hope is that this unsustainable state of affairs will resolve itself over time as "gold OA" becomes the norm. For now, responsible publishers have to build systems for differential charges in addition to waiver policies, which should be reserved for exceptional circumstances. A one size fits all author charge (APC) does not work across all disciplines, countries, funding status, and journal article types.
What comes around goes around: OA has not fulfilled the initial goal of a "free market" where authors chose the journal solely based on the value proposition. For one, publishing is so deeply embedded with research assessment that there never has been a choice purely based on the perceived value-added of a journal (unless the definition includes academic credit). In reality, OA has evolved from author choice as publishers are again locked in intense negotiations with consortia of research institutions to minimize costs of "Publish&Read deals." Large publishers wield more economic power and thus sideline nonprofit academic publishers, which often do not even get a seat at the negotiation table. Institutions and countries not embedded in publish and read consortia may end up effectively subsidizing large consortia with the economic clout to force charge discounts. Worse, many academic libraries are facing long-term budgetary pressures that prevent them from stepping into a key new role to support open science dissemination at academic institutions. The EU council recommends a shift to non-profit 'diamond OA' (that is centrally funded platforms free to readers and authors), without detailing coordinated strategies to achieve this in a scalable manner.

The future
Despite these challenges, the open availability of reliable research results is without any doubt essential for an efficient and effective publicly funded research process. At this pivotal moment in scientific communication, the community, funders, and governments have to work harder to agree on the primary goals of this conversion. The developments that are set in motion now will be hard to re-engineer later on. We need to derive a consensus on a clearly defined goal. In our view Open Access alone-even if it were to evolve into a sustainable process focused on quality-is not sufficient. At EMBO, we see that goal to be a deeply integrated collection of Open Science infrastructures, workflows, and policies that reaches well beyond the classical research paper, but that at the same time builds on existing processes and com-munication devices. These processes include constructive peer review, the narrativebased research paper, commentary and review for context and debate. Just as much, it includes structured community databases and data repositories, as well as preprint platforms. We must not disband effective mechanisms that have been tried and tested over many years and that underpin an incredibly dynamic era of biosciences research for the sake of often ill-defined alternatives. Instead, we need to aim to integrate these features into a more constructive process. For example, validated referee reports can be included far earlier with the preprint version of a journal article, adding context and quality assurance. Importantly, these refereed preprints can then also be realistically included in research assessment exercises. A refereed preprint can allow-in particular younger-researchers to become eligible for funding far earlier than the publication date of their research paper. This in turn depressurizes the race to publish papers, which can lead to more thorough revision and possibly a more reliable literature. EMBO now regards a refereed preprint on par with a research paper for application to its Fellowship Programme and is considering this for its Young Investigator applications. Another example: database curators have for years processed data entries independently of the relevant research paper. Ensuring that their expert curation also applies to the research paper will serve the scientific process and avoid duplication. Ultimately, curation and quality control executed by data centers at research institutions can be similarly integrated.
We must be open to experiment and invest in change where that change is promising to achieve our agreed goal. Today is a time of great opportunity for positive change. Lets embrace it! Reference Pariente N (2023) PLoS Biology at 20: Ain't no mountain high enough. PLoS Biol 21: