From the BRAIN Director: Sizing Up Science

photo of John Ngai

In my daily discussions with the BRAIN community, I hear the tension between small and big science: a proxy argument for exploring biology versus building tools. Individual labs do creative biology, so why fund large teams that churn out lots of data but aren’t testing hypotheses?

My answer is simple: Diseases don’t care about funding mechanisms.

Individual labs and large team projects are equally important for discovery. But given a fixed amount of funding, there is always a trade-off. Big science offers scale and efficiency, and small science promotes in-depth investigation; the two are key elements of a virtuous cycle.

Sometimes, large investments in team science are necessary to provide different types of knowledge at a scale simply unattainable by several small labs. Large, technology-focused programs (like two newly announced BRAIN transformative projects) serve thousands of individual labs within BRAIN and across neuroscience. Accessing large data sets created by these powerful (and expensive) tools brings items into focus for individuals to view and analyze biology in ways not previously possible.

One such tool is OpenScope, developed by the Allen Institute, which we have made available to BRAIN grantees. Sort of like the giant telescopes used by astronomers and physicists to look at far away stars and galaxies, OpenScope functions like a neurobiological observatory. This observatory offers efficiency of scale, speeding up what individual labs do not have the time or resources to do. A real trade-off, though, is that team science doesn’t always allow chasing down a new spark – that is why science works best in a range of sizes.

Using OpenScope, scientists might test an existing result at scale using live recordings. Or look at neuronal communication during behavioral tasks. Data collected are shared freely with neuroscientists and the community. This approach resonates with BRAIN’s commitment to open science – making experimentation, and its resultant data, as accessible as possible to any investigator with a good question, regardless of lab size, that will illuminate the biology underlying brain function.   

Beyond sharing sophisticated instrumentation, other opportunities to synergize small and big science include offering resources for generating and managing data resources for broad impact. Experiments in model systems, including fruit flies, zebrafish, worms, and others, can provide key insights for moving into larger, more complex systems: noted eloquently about 10 years ago by BRAIN 2025 Report contributors Dr. Cori Bargmann and Dr. Eve Marder.

For example, “Welcome to the Fly Brain!” welcomes visitors to the Drosophila Connectome. This resource was co-launched by the United Kingdom’s Wellcome Trust and the Howard Hughes Medical Institute. BRAIN is expanding its value by funding dissemination of FlyWire: an open-source game-like platform that is crowdsourcing the first complete wiring diagram of a Drosophila brain. To play the game (and to help us better understand brain circuits), players search for the right pieces and snap together beautifully colored 3D neurons. The “game board” is a fruit fly brain sliced and imaged by electron microscopy, in which artificial intelligence methods were used to identify pieces of neurons. This resource provides existing data to new explorers, toward understanding the fundamentals of brain development and function – ultimately across scales. Moreover, many BRAIN-funded and other neuroscientists are using Flywire to explore circuits of particular interest to them. This resource has also enabled basic science research to serve as a springboard for reconstructing complex mammalian brains.

Reinforcing the value of individual contributions to big science, the first-of-its-kind data set that makes up the Drosophila Connectome grew from the painstaking work of individual BRAIN-funded investigators. The first step was creating an efficient way to automate the process of mapping neural circuits, using a previously developed decentralized web interface that works sort of like Google Maps to sort and label terabyte-scale files of imaging data. That approach enabled BRAIN-funded researchers to predict how to connect synapses – almost like identifying linking pieces of a jigsaw puzzle – to generate an accurate connectivity map of the fly brain.

The OpenMind Consortium, also funded by BRAIN, offers another reason why we should resist the false dichotomy between individual labs and team projects. This open-source resource emerged when several BRAIN investigators working on implantable neurostimulation hardware platforms realized that they were running into the same hurdles and that lots of time and effort could be saved by working together. Their area of research, involving medical devices and clinical testing, has specific needs for launching clinical studies, and OpenMind offers tools and resources in a centralized space. These include a library of open-source software for home-based neural sensing paired with peripheral monitors and streamlined regulatory materials for FDA approval of investigational protocols. BRAIN-funded research served by OpenMind is an exciting area with near-term promise for improving quality of life for individuals with debilitating neurological conditions like movement disorders, epilepsy, and neuropsychiatric diseases.

Given the trade-offs I’ve discussed, we need to be strategic about when to invest in big science versus when to invest in innovation at a smaller scale. Mindful of the big picture, BRAIN takes a measured approach to develop opportunities that will benefit the entire field of neuroscience and those that provide chances for individual discovery to flourish using the latest and greatest tools. It’s all-hands-on-deck for untangling the complexity of big, small, and medium sized brains. We need creative ideas and novel hypotheses from diverse mindsets. We need technology engines that create new tools – and those that sort and analyze unfathomably large data sets – that allow us to ask and answer better questions. Moreover, we need open-source resources, some of which I’ve described here, to help distribute and democratize the fruits of our research to everyone. That’s essential toward finding treatments and cures as fast as possible for those in need.

With respect and gratitude,

John Ngai, Ph.D.
Director, NIH BRAIN Initiative