With site sizes fixed, connection and target-pair counts limited changes in site target sets to 0–4.05% in eight C. elegans brain maps

We tested whether connection counts and shared-target pair counts determine synaptic target sets in eight C. elegans brain maps. We also fixed the number of sites at each target count. Complete candidate enumeration and integer optimization gave attained limits on changes to unordered site lists. The two earliest maps had unique target-set multisets. The other maps allowed at most 0.43% to 4.05% of site records to be replaced. The adult limits were 80 of 3,422 records and 143 of 3,534. Removing site-size counts raised the adult limits to 3.04% and 4.67%. Releasing positive pair counts, while keeping connection counts, site sizes and absent pairs fixed, allowed 42.43% and 51.27% to change. Most exact sets with at least three targets had fixed multiplicities. The method gives a reconstruction audit before tests of larger synaptic groups. These bounds concern recorded target sets without locations. They do not identify physical sites or test physiological interactions. Pair counts contain additional site information beyond a neuron connection matrix. The package preserves code, source hashes, count intervals, feasible alternatives and reproduction instructions.

connectome · polyadic synapse · hypergraph reconstruction · shared targets · integer optimization · Caenorhabditis elegans · data compression