FlyWire — complete adult fly brain

FlyWire is the existence proof that a whole-brain connectome of a behaving animal is achievable. An adult female fly brain was imaged by ssTEM, segmented automatically, and then corrected by a community of proofreaders — thousands of hours of human labour — to 139,255 neurons and about 54.5 million synaptic connections, with a hierarchical annotation of classes, cell types and developmental units.

What it is good for. Graph analysis with a complete edge list. It is the only large dataset where “I have the whole network” is true, so it is the right place to learn null models: you can compare an observed motif count against a degree-preserving randomisation without worrying that your finding is an artefact of a truncated volume. It is also the clearest case study in what proofreading costs and who gets credit for it.

What it does not support. Generalisation to a second fly. This is one individual; the across-animal variability of the connectome is an open question, and the hemibrain and MANC datasets exist partly to address it. It is also structure only — no activity was recorded from this brain.