‘Open science’ practices like the public sharing of research data and protocols are becoming widespread: seen, for example, in the extensive use of preprint publications during the development of vaccines and treatments during the COVID-19 pandemic.
But in cutting-edge areas such as viral engineering and pathogen synthesis, the risk of accidental or deliberate misuse of biological research is also increasing.
Publishing their paper in Plos Biology last month, James Andrew Smith and Jonas Sandbrink from the University of Oxford call for greater collaboration between open science experts and biosecurity specialists to ensure that mechanisms are in place to ensure responsible research with maximal societal benefits.
“The move towards open science has involved a cultural shift related to conduct and sharing and research; a similar cultural shift may be required to encourage responsible research conduct and sharing to protect the life sciences from misuse,” they note.
Sensitive subjects: from nuclear physics to pathogen synthesis
Advances in immunology, virology and other biomedical domains have made great leaps forward in recent years. Meanwhile, the advantages of open science and sharing information as quickly as possible became evident during the COVID-19 pandemic.
But while developments in viral engineering are important in areas such as vaccine design and cancer therapy, these could also be applied to engineer pathogens with increased virulence or transmissibility. Deliberate release of such pathogens could result in a pandemic of unprecedented severity.
The researchers note that the conflict between transparency and security is not a new one: but one that needs to be reconsidered in the light of the advances being made in biologics.
“That certain information may cause harm and should not be publicly accessible has long been accepted in nuclear physics," note the researchers. "In 1946, the US Atomic Energy act turned all information on nuclear weapons into “restricted data” until formally declassified.
“Advances in molecular biology, including DNA synthesis and gene editing, are democratising science and lowering the barrier to the synthesis and engineering of biological agents.
“As biotechnology advances, we need to acknowledge that biological research, similarly to nuclear physics, may uncover information with security implications and consider the possibility that not all information should be made publicly available.”
Preprint publishing
The researchers outline several areas with potential threats - alongside potential solutions. One of the areas is preprint publishing: which offers a faster route to publishing research than traditional journals.
While major preprint servers have pledged not to post studies of pathogens that could cause harm, this was relaxed somewhat during the COVID-19 pandemic because of the need to disseminate knowledge about viral variants rapidly.
"Preprints may therefore remove the “gatekeeper” role that journals could play in mitigating risks from the publication of research with potential for misuse," note the researchers.
"Authors may select preprint servers that do not screen research. Unlike publishing in particular journals, there is little incentive to post to a particular preprint server, so little reason not to select one that will immediately post the article."
One solution here is that prominent preprint servers could work together to agree on a common policy.
"There are few prominent preprint servers in comparison to the number of journals, which may represent an opportunity for coordination. Relatively few parties would need to agree on a policy for it to be implemented universally, or at least across all of the major servers.
"If an article is flagged by at least one server as potentially concerning, other servers could agree not to post that article until it was appropriately peer reviewed. However, authors must be incentivised to post to those preprint servers with appropriate review processes, and funders, institutions, and possibly journals, rather than researchers, will likely need to encourage this."
Meanwhile, in the context of viral engineering, the researchers say that open code, data, and materials may increase the risk of the release of enhanced pathogens. Openly available machine learning models could reduce the amount of time needed in the laboratory and make pathogen engineering easier.
To prevent such misuse, mechanisms that ensure responsible access to relevant dangerous research materials need to be explored. In particular, to prevent the misuse of computational tools, controlling access to software and data may be necessary.
Creating the right mechanisms
While the authors explore several ways that checks could be introduced into the research process, the biggest impact will be made by addressing the overarching systems.
“Funders and institutions have an important role to play in improving biosecurity," they note. "Storing data in approved platforms, posting only to preprint servers with adequate review processes, and preregistration of research with high potential for misuse, for example, could be mandated or encouraged by funders or institutional oversight groups.
“While solutions are developed and implemented, clear policies should be in place for the communication of research outputs that involve safety or security risks.
“Greater investment in expertise related to biosecurity and biosafety will likely be important for realization of any proposals involving peer review for risk assessment or mitigation purposes."
Such investment in expertise could allow biosecurity experts to get involved in the peer review process, or in the early stages of research by making recommendations on how to conduct activities while minimizing risk.
At the very least, those working in the industry need to be aware that open science has both advantages and disadvantages, conclude the authors.
"More generally, consideration of downside risk of both open science and biological science appears to be neglected in comparison to its plausible magnitude. Education and outreach may help to increase awareness among relevant stakeholders.”
The full study (peer-reviewed, open-access) can be found here.