In the wake of COVID19, the world is scrambling to figure out how to continue the functions of normal life while confined to a much smaller world of couches and home offices.
In the wake of COVID19, the world is scrambling to figure out how to continue the functions of normal life while confined to a much smaller world of couches and home offices. Work, family and friend relationships, exercise, entertainment, and all the other elements of normalcy have been upended. The effort to maintain productivity (and, to be frank, sanity) has initiated a mass digital migration; a movement from life to e-life.
If you are a researcher you are likely dealing with your own unique micro-version of the predicament: how to continue working when you and your lab are self-isolated? In other words, how to replicate as many of the normal functions of a lab as possible in a digital environment; a movement from lab to e-lab.
Lab meetings have migrated to video-conferencing with relatively little friction; conferences have gone virtual astoundingly quickly; manuscripts are being drafted and figures created – perhaps a tad earlier than originally planned; grants and proposals are being written. These are, however, only a small portion of what researchers do, and leave a huge swath of scientific work untouched and in stasis.
Those most able to continue useful work are those who, for whatever reason, have access to as much of their work as possible from home. Pictures of researchers lugging multiple work computers home in order to continue data analysis were common in the early days of the crisis.
There is a subgroup of researchers, however, who can continue work without having to cram several computer towers into the trunk of a car: those who created an up-to-date digital, online representation of their research activities right up to the moment they had to be abandoned. These lucky few, who put in the time and effort to create a comprehensive e-lab prior to the crisis, are for the most part those who already commited to open science. This is, of course, no accident.
Open science is, at its most practical base, the effort to (1) create an accurate and well organized digital representation of research activities and (2) share that representation with as few barriers as possible so others can effectively check, replicate, rework, reuse, and remix findings. Even the concentration within open science on sharing physical materials – like cell lines, biospecimens, and model organisms – is to enable replication and follow-on research when sharing the information about those materials does not to allow others to generate them themselves due to the unique nature of the material or an inability to make materials locally.
Researchers who have incorporated open, online sharing of research resources throughout their work – sharing experimental designs and protocols, lab notebooks, data, software, etc – have, in effect, created a digital representation of a significant portion of their activities that can be accessed from anywhere with an internet connection.
Not to oversell things, open sharing of research resources does not mean that researchers and labs can continue functioning without a hitch in their experimental stride. Much of the actual physical work has necessarily halted. It does mean, however, that many of the inputs to and outputs of their research are available in a digital form; accessible from a couch, desk, or apartment window-sill. While the emphasis from open science advocates in the past has been on ensuring that these digital representations exist so they can be shared, an added advantage has now jumped to the fore: the ability to share your e-lab with your quarantined-self.
Labs that have, for example, stored experimental data immediately upon collection on open science platforms such as the Open Science Framework, Zenodo, Figshare, Synapse, OpenNeuro, and Dryad now have seamless access to the last bit collected before retreating into quarantine. Analyzing this data may be impossible if the requisite software only exists on a work computer, or you don’t have a license to use it at home; if, however, research software has been openly shared via a platform like Github then it can be accessed from a park bench (assuming a decent wifi signal).
If the collected data needs to be put into the context of how it was collected, perhaps to elicit input from other lab members about the best way to analyze it, there is no easier way than accessing the relevant protocols through protocols.io or an Open Lab Notebook. Need to know exactly what reagents or model organisms were used in an experiment? If one cannot go check the freezer, cage, or delivery invoice, then checking the relevant Research Resource Identifier (RRID) listed in your shared protocol or notebook is the next best (if not better) thing.
The list of unanticipated advantages continues. Perhaps you need to get input from a normally-down-the-hall colleague about how to improve your protocol for the future, or eliminate an artifact from some data, or iron out a persnickety bug in analysis software. In the lab you could invite them to pull up a chair as you took them through the issue. Now that is, obviously, impossible. If the protocol, data, or software were shared, however, you could email them the link (or, better yet, persistent identifier), open a videochat window, and collaborate from a distance.
One of the most heartening things I have seen while scrolling through my twitter-feed-turned-emotional-roller-coaster is that self-isolated researchers are engaging with openly shared research resources.
This is of course the case with COVID19 research, where openly shared software, data, manuscripts, and protocols relevant to the virus are being shared, collected, curated and adapted with blinding speed; but it goes farther than that. Every day I see someone in the neuroscience research community tweet something along the lines of “thank goodness for X piece of openly shared software/data that I can learn about and use from home!”
When people are cut off from their physical labs and don’t have an e-lab of their own, it is wonderful to see them being able to pick up the tools and resources shared by others. The lessons we are being forced to learn about the fluidity, dynamism, and capacity for collective creativity enabled by sharing research resources will serve us well in the future.
My sincere hope is that those working with these new found tools and resources take the open ethos to heart and openly reshare any modifications, derived data, and new insights with the research community. Doing so closes the loop, creating a community-based feedback mechanism that encourages further sharing in the future and, more importantly, uses the expertise of the world-wide research community to create ever better science.
Some of the people involved in shaping (see: arguing about) what open science is quail at the idea that open science has anything to do with equity and equality. They say that it is merely about replicability and transparency and that is it, harumph. What I hope this situation brings to light – as you swear at your institutional VPN for once again dropping you, or find yourself unable to download important papers because of paywalls your institutional account would normally breeze through – is that you are receiving a taste of what it is like outside the gilded ivory tower.
Such restrictions on access and use are what interested members of the public, or those at research institutes that cannot afford staggering subscription prices, or those in labs not well funded enough to pay for the licenses to fancy analysis software, face every day. The act of openly sharing a paper, a piece of software, a dataset, an equipment design, or whatever else is not just about replicability. Doing so is an act that expands the borders of science, allowing the participation of the previously excluded. It is an act of inclusion not in some abstract post-modern sense, but in the very real sense of choosing where and why to put the fences.
Beyond the fact that open sharing is useful to the isolated-sharer and their research community, shared research resources have taken on an unprecedented importance in the COVID19 crisis. Stories abound of how sharing genetic sequences and protein structures have led to the discovery of novel treatment targets, diagnostic testing protocols, and the initiation of clinical trials on a timescale that, in the pre-COVID19 era, would be unbelievable. Moreover, the rapid release of results via preprints has allowed the creation of competitions to apply machine learning to generate yet more insights. Though similar mass sharing has occurred during previous outbreaks (e.g. Zika and Ebola), the fact that COVID19’s impact has been global (and, not to put too fine a point on it, has impacted the global north) holds the hope that this time it will stick.
There is no reason that the same strategies cannot be applied to the ongoing crises in neurological disorders, antibiotic development, and the treatment of rare diseases, to name only a few. Indeed, the kind of massive online collaboration to solve issues around testing, creating ventilators and personal protective equipment, solving protein structures, and identifying novel treatment targets in the context of COVID19 is exactly what open science advocates have been dreaming of for years.
Open science has laid the groundwork for the creation of e-labs and for them to be linked up to enable global collaboration. The COVID19 crisis has, on so many levels, highlighted the importance and promise sharable, digital representations of research activity hold for the future of research. More people than ever are glimpsing the potential that e-labs interlinked into a global digital research ecosystem represents.
This bright future does not, however, come free of complications. Just as open science reveals the potential of sharing, it also reveals difficulties of coordination and consensus. If research is going to be conducted collaboratively on a global scale we need to make sure that shared resources are findable, accessible, interoperable, and reusable (e.g. FAIR). Doing so can be made easier through technology and sharing platforms, but what it requires most are education, communities collaboratively setting norms around standards and sharing practices, and a commitment to embed open science practices throughout the research lifecycle. The necessary changes in behaviour will require social and cultural innovation on the part of research communities at least as important as any technical advance.
Another big coordination problem has to do with eliminating research redundancies. Individual researchers within the global community need to coordinate so as to pursue different but complementary efforts. Doing so requires researchers to move past the stage of sharing resources in the hopes that those in their community will find it to engagement with that community. In the old, closed model such coordination was handled (albeit inefficiently) through journals and conferences where community members would be apprised of the work being done by others. The same task now must be undertaken in the digital world. Those of us interested in advancing open science are busily documenting how this is being done for COVID19 research in the hope that we can, after this is over, present an iron-clad case for how research communities can be rebuilt with digitally-mediated coordination planted deep in their hearts.
I would ask you to deeply consider how you can begin adopting open, collaborative, community based approaches in the future of your research, once the world is rightside-up again. Doing so isn’t just about some abstract idea of openness; it has real benefits for you, your lab, your research community, and the world. COVID is, it is true, producing never before seen destruction of research efforts from which many projects may never recover. Within this destruction, however, lay clues to how research can be remade better, more collaborative, more coordinated and, of course, more open.
About the author
Dylan Roskams-Edris joined The Neuro in November 2019 as the Open Science Alliance Officer for the Tanenbaum Open Science Institute. In his role, Dylan works to encourage the uptake of Open Science at neuroscience institutes across Canada. He has a background in neuroscience, health ethics, and intellectual property.
May 4 2020