Variable chemical wiring across 8 C. elegans brains did not predict more variable co-activity across animals; gap junctions in more brains gave stronger, and in moving worms more variable, co-activity

Question: do neuron pairs with wiring that varies between individuals also have co-activity that varies more between individual animals? We linked the eight Witvliet 2021 C. elegans connectomes (stable, developmentally dynamic and variable chemical connections; number of brains with a chemical synapse or a gap junction) to per-animal pair correlations in three whole-brain calcium data sets: Atanas 2023 (40 moving animals, 5,700 pairs), Baskoylu (27 moving animals) and Randi 2023 (112 immobilised, stimulated animals). Models controlled for adult synapse counts, per-neuron fixed effects, soma distance, ganglion and the mean. Inference used 2,000 QAP permutations and 1,000 animal bootstraps; the plan was written first. Result: pairs with only variable chemical connections did not have a larger across-animal spread than pairs with stable connections (Atanas log SD difference -0.004, 95% CI -0.073 to 0.082). Each additional brain with a gap junction increased the mean Fisher z (0.024, 0.022 and 0.009 per brain; all QAP p < 0.003) and, in the two moving data sets, also the across-animal SD (+3.3% and +3.2% per brain), not decreased it. Neuron-level wiring variability did not predict profile consistency. A post hoc split-half analysis put the extra spread mostly between animals with all labels, but not with high-confidence labels only. Limitations: wiring and activity come from different animals; few pairs have highly stereotyped gap junctions; label errors and state changes can explain part of the spread.

C. elegans · connectome variability · gap junctions · whole-brain calcium imaging · individual differences · structure-function · Witvliet 2021 · functional reproducibility