The goal of the International Brain Laboratory (IBL) is to unite many different member laboratories across the world to understand what happens across the mouse brain during decision-making. To do this, each lab uses the same standardized task and records from a different part of the mouse brain. Synthesizing these disparate datasets and formalizing the neural computations that support decision-making requires a high level of collaboration at every step in the research process.
IBL’s postdocs, graduate students and scientific staff have built a complex, grassroots network that helps us collaborate effectively, distribute tasks and resources, and produce resources for the greater scientific community. A full analysis of our approach can be found in this article, but below are a few key lessons for team-based neuroscience.
Develop common language to connect your team
A large part of successful collaboration comes down to speaking the same language, especially when different experts across diverse disciplines interact with each other. Shared terminology, colloquialisms and shorthand means that IBL members are able to discuss concepts across expertise, position or experience. For instance, using standard language to discuss a behavioral task helps connect disparate teams that work on data infrastructure, software issue tracking in Github, behavioral analysis or computational modeling. When someone says a mouse is ‘fully trained,’ we can be sure that other colleagues—from a diversity of experimental, theoretical, technical and administrative backgrounds—all understand what that means.
Make a space for sharing knowledge and building a community
Shared space is a given in most traditional laboratory environments but becomes a difficult concept once a collaboration expands beyond a department, university or country. The IBL includes scientists across 16 institutions and nine time zones, so IBL community spaces are almost entirely online. For virtual teams like ours, this has required several different online tools and information formats: Slack for real-time messaging, Google GSuite for documentation, Zoom for video conferencing, Github as our code repository and Datajoint as our custom experimental database. Initially, IBL’s early-career researchers spearheaded the use of these platforms thanks to the ever-changing, real-time demands of experiments. For instance, our Slack ‘troubleshooting’ channel allows us to gather and disperse know-how and solutions over the span of a few minutes. These days, this connectivity has been adopted by all team members; we regularly convene to ask questions, give advice, discuss policies or just informally chat with one another.
Keep a ‘flat’ organizational hierarchy
IBL’s organizational structure is directly inspired by the ATLAS collaboration at CERN, emphasizing horizontal relationships across colleagues in lieu of traditional vertical management hierarchies. A voting assembly ratifies important decisions, an executive body manages operations, and a flexible network enables specialized subgroups to take on particular tasks. Both PIs and early-career researchers can influence policy by directly contributing to team-wide actions or speaking to elected representatives with concerns or comments. This model has been very successful in diminishing the traditional hierarchy between trainees and PIs; traditionally postdocs or graduate students may have one or two mentors, but IBL early-career researchers establish rapport with a multitude of PIs who can offer scientific guidance and career advice.
Cultivate and connect information across different people and places
In principle, every piece of information in IBL is accessible by anyone at any time. This is fundamental for our organizational transparency, but the sheer amount of information is too large to navigate single-handedly. To stay organized, we delegate information across Working Groups (WGs) that each maintain specialized knowledge and connect to other groups by way of overlapping members (see figure, right). On top of WGs, ‘expert’ members across the collaboration informally steward information and apply it to new situations. This system of rapid responders keeps us nimble despite our size, but it can be vulnerable to drop-out when people leave the collaboration. Moreover, it can be complex to archive; an ongoing challenge for us is how to record our ‘organizational memory’ in a way that is accessible both short- and long-term.
Diversify efforts and encourage crossover of knowledge domains
Every large-scale effort that we have undertaken in IBL has required a rich diversity of contributions across different expertise. Conceptualizing experiments, designing protocols and visualizing the resulting data require both the formal knowledge stewarded by postdocs and students (theories, models, theses) and the contextual knowledge held by technical staff, (materials, instruments and protocols). Additionally, the crossover of theoretical and experimental expertise richly influences our approach to neuroscience; theories evolve to suit empirical data and experimental design is governed by computational models. There is ample opportunity for experts in one area to influence, contribute or simply learn from another, which broadens our scientific training and makes for better science.
Adhere to open-source principles and make public sharing the norm
From its beginning, IBL has committed to make all resources, data and tools publicly accessible; indeed, our first deployment of data and tools occurred earlier this year with the release of our platform behavioral paper. This strongly influences how we envision our relationship to other organizations and our responsibility to the general public. Our collective authorship practices and open-access mandates have occasionally uncovered obstacles in the traditional pipeline for sharing large-scale, collective scientific work; for example, many publishing platforms don’t easily accommodate the use of a consortium author like “International Brain Laboratory,” which we include on all of our publications. But our organizational heft has allowed us to start changing these procedures. We hope this open model will inform a new culture of scientific research that is friendly to broad and diverse contributions and that is completely open access by default.
Make the changes needed to push your team forward
Alongside scientific work, IBL researchers routinely advocate to improve research culture and respond to issues facing scientific practice both within and beyond the collaboration. Cultivating advocacy alongside science research may be one of the enduring touchstones of IBL’s collaborative model, providing next-generation scientists with agency to tackle issues like diversity and inclusion, career security or work-life balance. The next iteration of Big Science may buck tradition even further, pushing toward a fully collaborative, reproducible and open endeavor—and make room for an even greater diversity of scientists who come after us.
Lauren E. Wool is an IBL member and Marie Skłodowska-Curie postdoctoral fellow in the Cortexlab at UCL, studying sensorimotor encoding in cortical neurons in mice. She can be found tweeting about science at @laurenewool.