From sense to movement, without a break
A fruit fly's brain is roughly the size of a poppy seed. Inside that seed sit 166,691 neurons spanning the brain and nerve cord, every one of them now proofread, annotated and sorted into 11,691 types. The map was published on 3 September in Cell and Current Biology, after circulating first as a preprint. Beside it stands the female brain map unveiled in 2024, covering about 140,000 neurons. Stop on that for a moment: for the first time, an animal's entire nervous system, from sensory surface to muscle, sits in a single document.[1]
The comparison speaks in numbers. Placed side by side at synaptic resolution, the male and female brain connectomes yield 7,205 isomorphic, 114 dimorphic, 262 male-specific and 69 female-specific types. Those differences pile up in the higher brain centres, while the sensory and motor periphery stays largely isomorphic. The same smell, the same sound, the same touch enters both animals through the same door and parts company further in. Numerous circuit switches reroute sensory information into antagonistic circuits controlling opposing behaviours.[1]
The edges of the map
The resource comes from a single male fly. A small number of cells could not be reconstructed because of sample artefacts at the edge of the volume. Postsynaptic completion in the neuropils is 42 per cent, so a share of the counted contacts lands on neurons that have not been proofread. With no fully proofread female ventral nerve cord connectome available, sex comparisons for the nerve cord rely on prior literature. Gene expression data are incomplete, and the fruitless/doublesex annotations cover the central brain only.[1]
A wiring diagram tells you which cell touches which; it does not tell you what the touching computes. The value of this map today lies in the questions it lets others ask: which sensory neuron connects to the neurons controlling walking, where in a circuit a dimorphic type sits. A rival reading is available too: the clustering of dimorphic types in higher centres may be a trace left by a single specimen and incomplete gene expression data rather than a real concentration. What separates the two readings is a second fly drawn with the same care.[1]
The question of scale
The price of that poppy seed has been written down. Imaging ran at 8x8x8 nanometre isotropic resolution, produced 160 teravoxels over a volume of 0.082 cubic millimetres, and took 13 months on 7 systems. It detected 46 million presynapses and 312 million postsynaptic densities, with average precision and recall of 0.82 and 0.81. Proofreading took an estimated effort of 44 person-years. A human brain is millions of times larger than that seed, and closing the distance asks for a method that changes the scale, rather than more of the same patience.[1]
The near-term targets the teams have named are the brains of larval zebrafish and adult danionin fish. If the same imaging and proofreading pipeline is applied to a larval zebrafish central nervous system, a fully proofread vertebrate connectome could be published by the end of 2028; the number to watch is the proofreading effort in person-years that such a study reports. In Carlos Ribeiro's account, this work provides a technical roadmap for more ambitious projects such as those for mice and humans. A roadmap carries no promise: who walks it, on whose budget and for whose curiosity, is still open. What we know tonight is smaller and firmer. One nervous system, down to its finest point of contact, has been read all the way through.[1]