The scientific moment
Neuroscience has been chasing the same fundamental question for decades: How does a vertebrate brain produce the rich, flexible behavior of a living animal? Answering it both matters in its own right, and is core to making real progress against brain disorders that impact hundreds of millions of people, but the question outran the tools science had available — until now.
Two forces have converged to bring what was previously impossible within reach: a new generation of tools that let us observe whole biological systems with a completeness that wasn’t possible a decade ago, and AI capable of parsing data at scales no human scientist can. This is the moment when what was out of reach becomes tangible, and seizing the moment requires exactly what Janelia was built to be — a place designed for problems no university or company could take on, requiring capabilities, team scale, and time horizons beyond what any single lab could sustain.
Janelia’s next decade will be defined by two big bets. Bet One: Janelia’s biologists, AI scientists, tool-builders, engineers, theorists, experimentalists, and other specialists will target one of neuroscience’s most critical and fundamental goals: a mechanistic account of how a vertebrate brain produces complex behavior, from molecules, neurons, and circuits to the whole animal in motion. Bet Two: To accomplish this, we’re building a new mode of discovery — AI-in-the-Loop — in which scientists and AI work together as one continuous system. We’re building it both to reach Bet One, and as an end in itself: a new way of doing biology that we’ll release, along with the tools behind it, for laboratories everywhere.
Janelia Research Campus
STRUCTURE & PHILOSOPHY
Four HHMI Janelia Initiatives, one enterprise
Janelia is organized around a single grand challenge, and Janelia’s structure is built to make accomplishing this mission possible. The scientists who drive this work additionally benefit from, and contribute to, science carried out by the broader HHMI community — a fact that strengthens all aspects of HHMI discovery.
Biology pursues the research questions central to Janelia’s mission. Three additional Initiatives — AI@HHMI, Tools@HHMI, and Theory@HHMI — sit alongside Biology, developing the distinctive capabilities needed to answer those questions while also connecting those capabilities to science across the Institute. Their roadmaps are developed in relation to Biology’s scientific priorities, and to one another, adding up to four organizational units that operate as a single, integrated enterprise.
The program
AI-in-the-Loop Discovery
AI-in-the-Loop is a new mode of research in which scientists’ expertise and AI’s analytical power operate in a continuous loop, driving discovery at a speed and scale beyond anything biology’s seen before. Broadly, AI systems analyze complex data, build models, and help design tools, systems, and automated pipelines. Scientists and AI co-create hypotheses and experimental strategies. Scientists run experiments, make decisions, and collect high-dimensional data. AI learns from that large-scale data and the cycle repeats.
This is the methodology HHMI is committed to building. Where the underlying AI capabilities are mature, we use them. Where they are not, we advance them, working through AI@HHMIexternal link, opens in a new tab — our institutional commitment to artificial intelligence in biological research — and in close collaboration with Google and Anthropic.
Model Organisms
Fruit fly
Complete neural map
Janelia has decades of experience with the fruit fly. In 2020 it published the fly “hemibrain” connectome — then the largest and most detailed synaptic-level brain map ever produced — and in 2025 the first complete connectome of a male fruit fly’s entire central nervous system, brain and nerve cord together. With a brain wiring map of roughly 140,000 neurons, a precise genetic toolkit, and Janelia’s dedicated fly labs, the fruit fly is where we develop AI-in-the-Loop Discovery and our integrated molecular workflows.
Larval zebrafish
Whole-brain imaging
The larval zebrafish is the only vertebrate in which every neuron in the brain can be imaged at once during behavior, and it’s where we’ll build the whole-brain pipeline the adult Danionella platform will need. Our zebrafish work has already uncovered fundamental principles underlying sensorimotor processing and motor learning while pushing the boundaries of imaging technology. This is where whole-brain neuroscience becomes possible, and where we learn to scale it.
Danionella
Whole-organism vertebrate platform
Because Danionella remains transparent throughout its life, it allows Janelia scientists to work toward what’s never been possible in an adult vertebrate before: imaging the whole brain in a living animal during natural, complex behavior. That opens a path that larval systems alone can’t provide — toward recording, mapping, and ultimately understanding how biology becomes behavior. Over the coming decade, Danionella is where we pursue that understanding in full.
At Janelia, the metric for success isn’t credit, but adoption — a change in how science itself is done. The technologies, methods, and concepts Janelia produced in its first 20 years advance science across HHMI, and in labs around the world.
That practice will continue into our next chapter, and many of the capabilities we’ll produce and openly release as part of this program will turn out to be most valuable for areas of work that Janelia itself doesn’t pursue — and that’s by design. The biological discoveries we make matter in their own right, and the tools, methodologies, frameworks, and AI systems we build along the way are just as core to our mission as the ultimate goal.