One-synapse larval C. elegans edges with mirror support occurred more often in both adult brains after contact and repeat controls

We tested whether a mirror connection in one larva marks weak connections that occur in two adult brains. We used chemical synapse data from Witvliet et al. (2021). The sample had 1,065 one-synapse observations of 816 directed edges in three larvae, aged about 0, 5 and 8 hours. Both ends of each edge belonged to true left-right neuron pairs. Edges with a mirror connection occurred in both adults in 379 of 506 observations (74.9%). Edges without a mirror connection did so in 213 of 559 observations (38.1%). After controls for contact area, detection in other larvae, neuron degree and neuron type, the odds ratio was 2.27. The model mean probability difference was 15.8 percentage points. A fixed-penalty model gave a similar estimate. Its 95% range from joint resampling of neuron pairs was 2.2 to 28.9 points. Keeping one row per edge gave an odds ratio of 2.29. Adding mirror support to a model tested on excluded source pairs reduced Brier error by about 0.004. The resampling range included zero. Tests on excluded target pairs gave a similar result. The evidence supports an adjusted association in this fixed network. It does not show a large added prediction gain, synapse survival, or a causal effect. Only two adult brains supplied outcomes. The package contains numerical source extracts, source versions and hashes, code, figures, all results and repeat instructions.

Caenorhabditis elegans · developmental connectome · bilateral symmetry · weak connections · chemical synapses · contact area · cross-validation