Gap junction weight had a positive co-activity link after three reference population modes were removed in two worm cohorts

We tested whether shared population activity explains the link between gap junction weight and neuron pair co-activity in C. elegans. We used 40 Atanas, Kim recordings and 8 Dag recordings, with Cook anatomy and two adult Witvliet maps. Population modes were estimated from reference ROIs that excluded all tested neuron ROIs. The pair model controlled for neuron effects, behaviour, soma distance, neurite contact, chemical weight and shared chemical inputs and outputs. After three modes were removed, the mean Cook gap coefficient was 0.012943 in the 40-recording cohort and 0.008325 in the 8-recording cohort, in absolute Fisher correlation units per standard deviation of log gap weight. Animal bootstrap 95 percent intervals were 0.011221 to 0.014626 and 0.005385 to 0.010807. These means were 7.0 and 17.8 percent below the zero-mode values. The positive remaining association persisted with stricter labels, removal of same-class pairs, both adult maps and a window before heat delivery. Attenuation was less consistent across checks. The intervals condition on fixed anatomy and cells. Different animals, one laboratory, partial reference coverage, optical mixing and possible conditioning bias prevent a direct electrical-causation claim. Code, aggregate evidence, figures, hashes and retrieval instructions support reproduction. External records are kept outside the bundle.

C. elegans · gap junctions · population activity · structure and function · calcium imaging · partial correlation · connectome · reference ROIs · common drive