The Strength of Weak Ties

“It was the best of times; it was the worst of times,” Charles Dickens famously began in A Tale of Two Cities. So it was for scientific research in early 2020 as a number of forces came together to create a unique set of opportunities and challenges.

First, the COVID-19 pandemic itself. The disease was so contagious and so serious that physical, human-to-human proximity was canceled except for interactions essential to life. Laboratories closed; lecture theaters and libraries lay empty; people barely left their homes.

Second, the emergence of technology-mediated collaboration. Video conferencing became the new meeting space; social media were repurposed for exchanging real-time information and ideas; and digital architects put their skills to building bespoke platforms.

Third, the scientific world united around a common purpose: generating the evidence base that would end the pandemic. Goodwill and reciprocity ruled. We forgot about academic league tables, promotion bottlenecks, h-indices, or longstanding rivalries. We switched gear from competing to collaborating. We pooled our data and our expertise for the good of humanity (and, perhaps, with a view to saving ourselves and our loved ones). And not to be overlooked, the red tape of research governance was cut. Our institutions and funders allowed us—indeed, required us—to divert our funds, equipment, and brainpower to the only work that now mattered. Journal paywalls were torn down. It became possible to build best teams from across the world, to get fast-track ethics approval within hours rather than weeks, to generate and test bold hypotheses, to publish almost instantly, and to replicate studies quickly when the science required it. The downside, of course, was the haystack of preprints that nobody had time to peer-review, but that’s a subject for another day.

In “How to Catalyze a Collaboration” (Issues, Summer 2023), Annamaria Carusi, Laure-Alix Clerbaux, and Clemens Wittwehr describe one initiative that emerged from those strange, wonderful, and terrifying times. The project, dubbed CIAO—short for Modelling COVID-19 Using the Adverse Outcome Pathway Framework—happened because a handful of toxicologists and virologists came together, on a platform designed for exchanging pictures of kittens, to solve an urgent puzzle. Through 280-character tweets and judiciously pitched hashtags, they began to learn each other’s language, reasoned collectively and abductively, and brought in others with different skills as the initiative unfolded.

Online collaborative groups need two things to thrive: a strong sense of common purpose, and a tight central administration (to do the inevitable paperwork, for example). In addition, as the sociologist Mark Granovetter has observed, such groups offer “the strength of weak ties”—people we hardly know are often more useful to us than people we are close to (because we already have too much in common with the latter). An online network tends to operate both through weak ties (the “town square,” where scientists from different backgrounds get to know each other a bit better) and through stronger ties (the “clique,” where scientists who find they have a lot in common peel off to share data and write a paper together).

The result, Carusi and her colleagues say, was 11 peer-reviewed papers and explanation of some scientific mysteries—such as why people with COVID-19 lose their sense of smell. Congratulations to the CIAO team for making the best of the “worst of times.”

Trisha Greenhalgh

Annamaria Carusi, Laure-Alix Clerbaux, and Clemens Wittwehr candidly and openly describe their technical and soft-skill experiences in fostering a global collaboration to address COVID-19 during the pandemic, drawing from an existing Adverse Outcome Pathway approach developed within the Organisation for Economic Co-Operation and Development. The collaborative, nicknamed CIAO (by the Italian members who would like to say, “Ciao COVID!”), found much-needed structure in the integrative construct of adverse outcome pathways (AOPs), or structured representations of biological events. In particular, one tool the researchers adopted—the AOP-Wiki—provided an increasingly agile web-based application that offered contributors a place and space to work on project tasks regardless of time zone. In addition, the AOP structure and AOP-Wiki both have predefined (and globally accepted) standards that obviate the need for semantics debates.

Yet the technical challenges were meager compared with the social challenges of people “being human” and the practical challenges of bringing people together when the world was essentially closed and physical interactions very limited. Carusi, Clerbaux, Wittwehr and their colleagues stepped up during this time of crisis by exercising not only scientific ingenuity but also social and emotional intelligence. They helped bring about, in essence, a paradigm shift. There was no choice but to abandon traditional in-person approaches that were no longer feasible and to embrace virtual and web-based applications. Collaborative leads leveraged their own social networks in virtual space to rapidly make connections that critically helped the AOP framework become the proverbial (and virtual) “campfire” for bringing the collaborative together.

Importantly, this work was not constrained by geography or language. For instance, the AOP-Wiki allowed for asynchronous project management by people living across 20 countries, breaking down language barriers through incorporation of globally accepted lingua franca for documenting and reporting COVID-19 biological pathways. Data entered into the AOP-Wiki were controlled using globally accepted standards and data management practices, such as controlled data extraction fields, vocabularies and ontologies, and FAIR (findable, accessible, interoperable, and reusable) data standards. These ingredients provided the collaborative its perfect campfire for cooking up COVID-19 pathways. All that was needed were the “enzymes” to get it all digested. That’s where the authors stepped in, gently “simmering” the collaborative toward a banquet of digitally documented COVID-19 web-based applications.

The collaborative’s resultant work was the personification of the adage when there is a will, there is a way. The group’s way was greatly facilitated by a willingness to accept and leverage new(er) technology and methods (i.e., web applications and digital data) that enable humans—and their computers—flexibility and efficiency across the globe. Novel virtual/digital models enhanced the collaborative’s experience. Notably, the collaborative’s acceptance and use of the AOP framework and AOP-Wiki’s data management interface means the COVID-19 AOPs are digitally documented, readable by both machines and humans, and globally accessible. The AOP framework has not only catalyzed the collaboration, but prospectively catalyzes the ability to use generative artificial intelligence to find and refine additional data with similar characteristics. This means the COVID-19 AOPs may evolve with the virus, updating over time as new information is automatically ingested.

Michelle Angrish