MaleCNS v1.0 is reported as a structural map of an adult male Drosophila melanogaster brain and ventral nerve cord, containing more than 166,000 reconstructed neurons and approximately 11,691 cell types. Google Research reported the project on September 3, 2026, with HHMI Janelia Research Campus among the research collaborators.
That map has also become an unusually vivid computational-neuroscience test bed. Reported software demonstrations connect a digital fly-brain model to Doom and Super Mario 64. The important detail is easy to lose in a punchy headline: no living fly is shown playing either game.
The adult male fruit fly’s central nervous system is now mapped
A connectome is a reconstruction of neurons and the synapses linking them—a wiring atlas for a nervous system. MaleCNS covers both the brain and the ventral nerve cord (VNC), the long structure that carries neural connections through the body. Project materials report more than 166,000 reconstructed neurons and approximately 11,691 cell types.
The distinction matters because “brain map” can sound narrower than the project actually is. MaleCNS is a central-nervous-system map, not just a picture of the brain. Its reconstruction combines millions of two-dimensional images into three-dimensional neural shapes using computational and AI-assisted methods, followed by human analysis and classification.
The official visualization below rotates that mapped structure and labels its approximately 11.7K cell types. It is a view of the connectome’s anatomy—not a recording of a fly thinking in real time.
What a connectome actually maps
A connectome tells researchers which neural structures connect to which others. That is enormously useful, but it is not the same as watching electrical activity unfold moment by moment. The map provides the structural possibilities through which information may travel; experiments are still needed to connect that wiring with function and behavior.
The earlier FlyWire project shows why the distinction is valuable. Its 2024 adult fruit fly brain connectome contained nearly 140,000 neurons, more than 50 million synapses and over 8,400 annotated cell types. Researchers used it to investigate circuits associated with vision, taste, movement and other behaviors.
In other words, the map is closer to a road atlas than to a live traffic camera. It can show the routes. It does not, by itself, show every signal moving along them.
MaleCNS and FlyWire are related, but not interchangeable
Both projects map Drosophila neural wiring, but they describe different mapped subjects and should not be treated as one release or one specimen.
| Project | Mapped subject | Reported scale | What it helps researchers study |
| FlyWire | Adult fruit fly brain | Nearly 140,000 neurons; more than 50 million synapses; over 8,400 cell types | Brain-wide wiring and circuit analysis, including sensory and motor pathways |
| MaleCNS v1.0 | Adult male fruit fly brain and ventral nerve cord | More than 166,000 reconstructed neurons; approximately 11,691 cell types | Central-nervous-system structure and comparisons of male and female connectivity |
The male–female comparison is one of MaleCNS’s useful angles. Some neurons occur in both sexes but connect to different neighboring neurons. That gives researchers a way to examine sex-dependent wiring instead of treating every nervous system as a generic interchangeable circuit board. Biology, as usual, declines to be tidy.
How the digital Doom and Super Mario 64 experiments work
The reported demonstrations use software wrapped around a digital connectome model. Game frames provide input to simulated sensory neurons. Activity inside the model is then translated into game controls.
The reported Doom setup also describes a reinforcement signal: after damage, two PPL101 dopamine cells receive a simulated stimulus. That is an architectural detail of the software experiment, not evidence that a biological fly felt danger, understood the game or learned to survive.
The Super Mario 64 demonstration follows the same broad idea: a digital model is coupled to a game environment so that simulated neural activity can produce control signals. The model is “playing” through an input-output loop, not through a physical brain connected to a computer.
That difference is the whole story. A simulation can instantiate a wiring model in software without turning the model into a conscious animal—or proving that the original animal would behave identically.
Why the map matters beyond the game demos
The games are memorable because they make an abstract connectome visible: pixels go in, model activity is calculated, and controls come out. But the scientific value is broader and less flashy.
Researchers can use the wiring map to:
- trace pathways from sensory inputs toward motor outputs;
- compare the connections of corresponding neurons between males and females;
- investigate circuits involved in vision, taste, movement and stopping behavior;
- build computational models that test hypotheses about neural circuits; and
- explore Drosophila anatomy, connectivity and 3D data through resources such as Virtual Fly Brain.
The adult fly is small, but its wiring is not trivial. The FlyWire reconstruction contains approximately 149 meters of reconstructed wiring packed into a brain about the size of a grain of sand. Around 85% of its neurons are intrinsic to the brain, meaning they synapse only with other brain neurons.
Those details turn the map from a colorful scientific render into a usable research scaffold. Scientists can ask more precise questions because they have a much more complete structural reference to work from.
What the demonstrations do not prove
A digital fly-brain model connected to a game does not show that:
- a living fly played Doom or Super Mario 64;
- the model is conscious or sentient;
- the model successfully learned to survive;
- the digital connectome is biologically equivalent to a living fly; or
- a physical brain-computer interface was created.
A connectome is a structural map. It can narrow the possibilities for how a circuit works, but it does not provide a complete account of neural dynamics or consciousness. That is why the most interesting result is not “a fly ran Doom.” It is that researchers now have a detailed wiring scaffold that can support increasingly concrete computational experiments.
The bottom line: MaleCNS makes the adult male fruit fly’s brain and ventral nerve cord available as a large, structured model for neuroscience. The game demonstrations are a striking way to exercise that model, but the real breakthrough is the map—and knowing exactly what a map can, and cannot, tell us.
