While states have closed their borders in response to the coronavirus outbreak, science has opened up in a unique way. Researchers have been openly sharing their outputs and making research available across disciplines, publishers have broken down their paywalls, and new ways of creating and disseminating scientific knowledge have been developed. This blog article provides an overview of the manifestations and features of open science over the past few months.
(Tämä artikkeli on saatavilla myös suomeksi.)
”Research on the virus is advancing at an unprecedented rate, and researchers around the world are also sharing their research results in a whole new way. It is possible to develop a drug for the virus and development work is being done in numerous laboratories,” says Tarja Sironen, Associate Professor at the Department of Virology, University of Helsinki, in Yleisradio’s interview on March 2020.
In response to the spread of coronavirus (SARS-CoV-2) around the world and the illness caused by the virus (coronavirus disease, COVID-19), cities are locked down and states have closed their borders. At the same time, researchers have embarked on large-scale global collaborations, in which the openness of science and open science practices have been given a central role.
This blog article provides an overview of the manifestations of open science during a pandemic. Its themes are:
- Researchers share their outputs
- Collaboration across disciplines
- Infrastructure supports open science
- The role of preprints in scientific dissemination
- Publishers are breaking down their paywalls
- What is crucial information during coronavirus pandemic?
- Openness and transparency in decision-making
- The impacts of open science
- Open science requires responsibility
- Open science after the pandemic?
There has been a huge number of articles on the subject and this article provides only a concise, exemplary overview of the subject. If you come up with suggestions for additions or topics for a follow-up article, please comment in the message field or send an e-mail (email@example.com).
University of Helsinki researchers and COVID-19
Researchers openly share their research results and data
Researchers at the University of Helsinki, researchers in Finland and researchers around the world have in recent months participated in the COVID-19 study in a way that has been called unique and unparallelled. This has also boosted research, as Jeremy Farrar, director of the British Wellcome Trust, said as early as the end of February: ”An unprecedented amount of knowledge has been generated in 6 weeks.”
What is particularly relevant about coronavirus research is the extent to which researchers have openly shared research outputs – both the results and the underlying research data – for the benefit of others. The pandemic has highlighted the benefits of open science and has also helped to raise awareness of open science practices and the preconditions for openness, such as the FAIR principles for data sharing (see more at Fairdata.fi).
Sharing knowledge to promote science is, of course, a basic premise of science – ”just science” – but in practice there have been a variety of obstacles to open access, such as competition in the merit system and a costly publishing culture. Indeed, one aspect of open science is that it seeks to change culture (that is, practices) and bring science back to its roots: ”open science is just science done right”. With the pandemic, the game of science seems to have given way, at least for the time being, to the pursuit of the common good and better knowledge.
Collaboration across disciplines
Like in scientific publishing, there are also cultural rigidities in interdisciplinary collaboration. The discipline-based approach is supported by the way research is valued and resourced. According to Benedikt Fecher, the head of the research programme ”Knowledge Dimension” at the Humboldt Institute for Internet and Society, research on COVID-19 is characterized by a problem-oriented approach that has lowered the boundaries related to the disciplines. Why? Because no single discipline can comprehensively understand a complex research problem:
”Especially in such acute crisis situations, it is important for researchers to collaborate across disciplines and to work together with doctors, nurses, politicians and public administrators. What vaccines might be developed? How can we ensure the continued provision of medical care? How can we avoid an economic collapse? None of these questions can be answered by a single discipline. – – What strikes me is that, when things get difficult, researchers step up to the game. At the moment, we are seeing researchers who are embracing complexity and overcoming institutional complex avoidance mechanisms. – – Researchers lead by example. They are ignoring artificial disciplinary boundaries, skipping the journal publication system and making themselves heard whenever and however they can.” (Fecher 17.3.2020)
Research collaboration is now taking place across the scientific field, and new projects have emerged at a rapid pace, such as collaborations on research data or various hackathons (see also ”Finnish researchers join forces to investigate the airborne transmission of coronavirus”).
Infrastructure now better supports open science
The current pandemic is exceptional in scale, which explains the collaboration of researchers: it connects researchers from all over the world. Additionally, compared to previous epidemics in the 21st century, such as SARS at the beginning of the millennium, the conditions for open science are also better than before. Digitalization and research infrastructure better support the sharing of research results and data. In Finland, the main provider of infrastructure services is CSC – IT Center for Science, which has reserved resources for activities that support COVID-19 research.
The importance of open science infrastructure, such as open access publishing channels and data repositories, has grown over the past few months. The GenBank database provided by the NCBI and the China National Genomics Data Center have been key venues in opening up research data. Repositories, such as PubMed Central and GisAid, and preprint services, such as BioRxiv and MedRxiv, have also made a significant impact on sharing scientific knowledge and promoting research. Open data and information related to the COVID-19 research have also been compiled on a variety of platforms:
- Virus Outbreak Data Network seeks to provide all coronavirus-related data in accordance with FAIR principles.
- CSSEGISandData is a data repository at Github, operated by the Johns Hopkins University.
- COVID-19 Open Research Dataset (CORD-19) is a resource of scholarly articles about COVID-19 and related coronaviruses, updated by the Allen Institute.
- WHO is gathering the latest multilingual scientific findings and knowledge on COVID-19.
Overviews of the spread of COVID-19 spread have been compiled into the real-time reporting view of Johns Hopkins University and the daily reports of the World Health Organization (WHO). However, in the case of WHO, the organization’s restrictive licensing policy has also been highlighted. Access rights determine how information can be used, and the CC BY-NC-SA licences used by the WHO do not optimally support open data sharing.
”The NC (non-commercial) condition is a problem because defining non-commercial use is difficult. For example, articles should not be shared on commercial or advertising funded platforms, such as ResearchGate or paid education. Thus, information is not spreading as effectively as one might hope. When this is combined with the SA (share alike) condition, all derivative works must be distributed under the same license in the future. Non-profit organizations usually share their material under a CC BY licence, and the European Commission has also adopted CC BY and CC0 for sharing information,” says Soile Manninen, an information specialist who specializes in copyright at the Helsinki University Library.
Preprints accelerating scientific communication
The coronavirus has brought preprint articles – that is, a manuscript version of an article before peer-review – to wider awareness. With the evaluation and processing of a research article taking up to months, preprints are intended for fast and open information sharing, which is very important in a pandemic:
”[P]reprints sped up data dissemination during the Zika epidemic of 2015–16 and the West African Ebola outbreak of 2014–16. Most of the preprints appeared more than 100 days before a journal published the work. But overall, less than 5% of the journal articles about the two epidemics were first posted as a preprint. The COVID-19 outbreak has broken that mold.” (Science 26.2.2020)
The purpose of the preprints is to make the research results quickly available to the scientific community, after which they are subject to conventional scientific evaluation. Preprint services also have their own, albeit limited, review mechanisms to ensure quality and prevent abuse. During the coronavirus outbreak, preprint reviewing channels have also been established, such as the Outbreak Science Rapid PREreview. Because the value of preprints is related not only to openness but also to speed, they are linked to competition between researchers. That is why critical consideration is needed.
”Preprints also involve risks – they aim to ensure the priority of one’s own research results on the science timeline, which can make it very tempting for researchers to publish their preliminary results prematurely and regardless of uncertainties”, says Eva Isaksson, an information specializing in bibliometrics and a member of the arXiv user community.
More than 2300 articles on COVID-19 have been uploaded to the bioRxiv and medRxiv preprint repositories – and the number is growing by the tens every day. Due to the large number of research papers, bioRxiv and medRxiv have added banners to the top of their pages to remind those less familiar with the scientific process about the role of preprint articles in scientific communication.
Publishers break down their paywalls
On 30 January, the World Health Organization (WHO) declared ”a public health emergency of international concern over the global outbreak of novel coronavirus”. The next day, the publishers of scientific journals published a petition to the scientific community calling for the open publication of research outputs – data, publications, preprints – ”as rapidly and widely as possible”. More than hundred signatories to the statements will also commit themselves to similar activities in the future.
On March 11, the WHO reported that COVID-19 ”can be characterized as a pandemic”. A couple of days later, chief science advisors of the twelve countries appealed to scientific publishers to open up all coronavirus-related research, including data science methods (e.g. machine-readable formats, licensing, etc.). On March 16, following a pandemic declaration and a petition from science advisors, more than 30 publishers said they would open up their COVID-19-related publications and data. There was even some competition for openness, for example, STM publishers declared they will also streamline the reviewing process and provide analytical tools for free use.
These have all been natural and important steps of great worth in the global crisis, and have made even major publishers such as Elsevier, Springer Nature, Taylor & Francis, and Wiley momentarily supporters of open science. However, critics with good memory have pointed out that large publishers did not show up to help during previous epidemics, such as the Ebola or Zika virus, even though similar statements were issued at the time. And the broader issue of paywalls also exists.
How much research should be opened?
As publishers open up their COVID-19 publications, the question has also been raised as to what kind of research should be openly available in this situation. Researcher and director of punctum books Vincent W.J. van Gerven Oei addressed the issue in his blog post:
”The Wellcome Trust statement raises the fascinating question concerning what type of research is exactly “relevant” to the outbreak. As the outbreak of the COVID-19 pandemic is not only tied to the DNA of the SARS-CoV-2 virus, its protein structures, and the way [it] interacts with the human body, but also the field of medicine, and therefore also healthcare, and healthcare funding, and health education, and thus also much broader questions of state organization, economic structures, educational resources – in brief, all the ways in which humans have ordered the world. If we want to come to a full understanding of the outbreak, all peer-reviewed research in medical, STEM, social science, and the humanities is potentially ’relevant’ and should therefore be made open. But that is certainly not how Elsevier c.s. see it.” (van Gerven Oei 19.3.2020)
In the editorial in Geoscience Communication, the point of view on pandemic was expanded to include climate research, geosciences, hydrogeology and geochemistry. The data perspective of behavioural science has also been highlighted, as human behaviour has not been the least important factor in the spread of a pandemic.
Van Gerven Oei also raises a tricky question: for how long should COVID-19 research be openly available? For as long as necessary or as long as the world’s attention is on the spread of the disease?
Openness and transparency in decision-making
States have taken exceptional action to curb the corona pandemic. At the same time, the question has arisen about the openness and transparency of decision-making: What information are exceptional measures based on? In the case of China, there have been reports of delays in information at an early stage, in the United States information on the COVID-19 situation has been hidden, in Finland estimates of the disease’s progress have been criticized – and the Swedish action has been marvelled at. Openness and transparency in decision-making have also been highlighted from a scientific perspective. An editorial in Nature looked at the UK’s different strategy:
”[T]he evidence behind this approach was not revealed. Not unexpectedly, the approach was questioned by scientists, including epidemiologists and other infectious-disease specialists, and is no longer part of UK policy. Researchers understand that sudden changes in policy will be necessary in a rapidly evolving situation in which there are many unknowns. But governments risk losing their trust by announcing those policies before the underlying data, models and assumptions have been released. – – Politicians and their science advisers need to get with the times and embrace open research.” (Nature 17.3.2020)
An interesting practical example of openness was provided by Professor Neil Ferguson of Imperial College London. In his tweet, he explained the code behind the UK strategy (in the beginning) and the figures describing the progress of the pandemic. At the same time, he revealed that the code had not been described or documented, even though 13 years had passed since its development. Nor had the code been opened up, even though it was publicly funded. These – probably conventional – practices seemed problematic in the light of the acute crisis.
The impacts and outcomes of open science
During the COVID-19 pandemic, the openness of science has shown its importance and benefits in practice. The basic mantra of open science in terms of speed, price, access, quality, efficiency and impact seems to hold up quite well in exceptional circumstances:
”Open science aims to share scientific information almost in real time online and free of charge to the user by providing free access to research publications and open access to scientific data. Open Science increases the quality and impact of science by placing all information in one, easy access area. This can make science more efficient through better sharing of resources, more reliable through better verification and more responsive to society’s needs.” (EOSC Secretariat 23.3.2020)
Cooperation between the scientific community and the open sharing of research outputs has meant better research, including the verification of results and the detection of errors. The genome of the virus was published as early as January and made freely available on GenBank. When the virus genome was sequenced, the development of diagnostic tests and research into finding a vaccine began. Although it takes months to develop a vaccine, research has progressed ”with record-setting speed”.
In addition to success stories, it has also been recalled that, from the point of view of the accumulation of scientific knowledge, failed experiments and methods are also part of the openness of science, as they can increase knowledge of a complex phenomenon:
”To speed up research, it’s also crucial to share things that don’t work, [Dave] O’Connor [Wisconsin National Primate Research Center researcher] says – for instance, when experiments show an animal species can’t be infected with the novel virus. ’That’s important information that is not typically shared through traditional channels,’ he says – – .” (Science 26.2.2020)
Openness requires responsibility
Research infrastructures may better support the implementation of open science, but openness also requires responsibility as well as following good scientific practice and procedures. The rules of the game of openness and everyday practices – such as describing and referring to data – must be clear, because also the rules of competition still apply:
”When information can be misused, skewed or misinterpreted at the global level so quickly, we also need scientists and the public to treat open science with great care and responsibility. Without care and responsibility, there is a danger that open science can contribute to the spread of misinformation.” (EOSC Secretariat 23.3.2020)
And it is not just about researchers. For example, the question of the nature of preprints has arisen during the COVID-19 pandemic. In addition to researchers, scientific information is also used by other actors who may not be familiar with the practices of scientific communication. A well-known example is a preprint article published in bioRxiv, which speculated that the DNA of the SARS-CoV-2 virus had similarities to the HI virus that causes AIDS. The preprint article received dozens of critical comments at bioRxiv within a couple of days, and the research community quickly rejected the results – and this also led to a a peer-reviewed study. However, a widely spread false research result does not easily disappear from the world: according to the Altmetric online service, the preprint article now removed from bioRxiv has collected more than 20 000 more mentions than the peer-reviewed article rejecting its results. During the COVID-19 pandemic, moreover, rumours and conspiracy theories have been spread on purpose. The degree of such fabrication in scientific communication led researchers to draft a petition to curb it.
Is science changing and if so, in what direction?
The importance of research infrastructures supporting open science has become clear with the pandemic. The COVID-19 process has highlighted funders, actors and projects that support the transition to lasting openness in science. But as noted already, during the Plan S process a year ago, the open science infrastructure was not complete. The infrastructure perspective, however, has been taken into account in the strategy of the University of Helsinki.
One of the largest open science infrastructure projects in recent years has been the European Open Science Cloud (EOSC), which aims to open up science by developing services and, in particular, interoperability between services:
”The goal of EOSC is to open up all scientific data and publications and combine the results to drive new discoveries and tackle key societal challenges. EOSC is an ideal tool to respond to public emergencies such as the COVID-19 virus by opening up scientific data on the virus and sharing live on-the-ground data on the spread of the virus.” (EOSC Secretariat 23.3.2020)
What about the scientific culture – the practices and actions related to scientific communication and publishing? If and when the openness of scientific knowledge and the practices of open science have proved their worth in a crisis situation, will there be something permanent about them? This has been asked. Many researchers and students locked down in their homes may also have woken up to the fact that the vast majority of scientific information is not openly available (even though open access channels can be found).
Time will tell. But until then, one can, like researcher Benedikt Fecher, envision how a pandemic will contribute to the creation of a new and better research culture:
”And as serious as the situation is, when the crisis is over and its history is being written, COVID-19 could go down as a turning point for open science and purposeful science communication. And it could awaken our dormant institutions from their slumber and prompt a long overdue change. The questions that spring to mind are the following: how can we strengthen problem-oriented, interdisciplinary research? What could open and innovative publication forums that go beyond disciplinary closed access journals look like? How can we design effective forms of evaluation that interfere as little as possible with researchers’ work and the natural flow of knowledge? How can we foster meaningful science communication? What can research funding agencies, economic policy and science management do to ensure that complexity is reflected in the social order of science?” (Fecher 17.3.2020)