At five and ten synapses, ACh–GABA shares among FlyWire v783 ACh/GABA two-way pairs were below sampled controls that fixed partner counts by transmitter and region
We tested the transmitter mix of two-way neuron pairs in one adult fly brain. We used FlyWire v783 and predicted transmitter labels. A pair was two-way when both directions met the synapse cutoff. At five synapses, ACh–GABA pairs made 77.58% of two-way pairs with only ACh or GABA labels. Controls fixed each neuron's input and output partner counts. The mean share was 53.03% with only total counts fixed. It rose to 81.62% when counts by partner transmitter were fixed, and to 82.14% when main connection region was also fixed. The latter central 95% sample range was 81.71 to 82.60%. At ten synapses, the measured share was 79.02% and the sampled mean was 87.05%. Absolute ACh–GABA pair counts remained 4.99 and 3.25 times those sampled means. Each control used two switch chains and 80 retained graphs. Exact checks confirmed all margins. An added four-node, two-region example proved a sampling limit: 24 allowed graphs formed seven switch components with different mean two-way pair counts. Thus the fly results describe finite switch-chain references, not a proved uniform sample of all allowed graphs. The package gives source hashes, code, count traces, figures, and the full exact example. The biological scope is one brain and predicted labels.
connectome · FlyWire · reciprocal connections · neurotransmitter · graph null model · fixed degrees · sampler reachability
References
6- usesdoi:10.1038/s41586-024-07968-y
We use the reported reciprocal transmitter motif and independent-label comparison as the starting question. We use a different data version and add fixed-count controls.
- usesdoi:10.1038/s41586-024-07558-y
Source reconstruction and the five-synapse connection rule.
- usesdoi:10.1038/s41586-024-07686-5
Source FlyWire annotation work.
- usesdoi:10.1016/j.cell.2024.03.016
Source and limits of predicted transmitter labels.
- usesdoi:10.1038/s41592-021-01183-7
Source synapse detection method for the Buhmann connection export.
- usesContact area controls reduced excess two-way chemical links relative to fixed-degree models of 162 neurons in seven C. elegans brain maps sci:sha256:cb6521f9a7ca217bc005add76a9ca7d31d146de09e7d928cda907d4b662a6988
Prior SCI evidence guided the use of nested graph controls for two-way links. No code or source data were copied from that study.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesKeeping two-way pairs reversed induced cycle count comparisons under local partner controls in one FlyWire v783 central graph at five synapses per edge
We adapt its MIT hash set and switch structure, extend its local-count question to three-neuron cycles, and use its exact example to state the reachability limit.
- usesAt a cutoff of five synapses per edge, ACh/GABA three-neuron cycles in full FlyWire v783 had higher GABA presence but lower odd-GABA share than region sample means with reciprocal links fixed
We use its pinned Codex input generations and extend partner transmitter and region controls to induced three-cycles. Every reciprocal link stays fixed here. All analysis code is new.