Gap-junction coefficients were larger for slow than fast calcium correlations in two worm samples, but normalized gains did not repeat

We tested whether annotated gap-junction pairs had a larger adjusted correlation for slow than fast calcium activity. We linked a fixed Cook hermaphrodite map to identified neurons in 40 Atanas/Kim animals and eight separate Dag animals. Slow was 0.01-0.05 Hz; fast was 0.15-0.40 Hz. Per-animal regressions controlled chemical wiring, atlas distance, mirror identity and additive neuron effects. The mean slow-minus-fast Fisher coefficient was 0.1131 (nominal 95% animal-bootstrap interval 0.0952-0.1308) and 0.0634 (0.0344-0.0943). All 48 animal differences were positive. All fixed sensitivity tests had positive cohort intervals. However, the raw-outcome partial correlation difference was 0.0135 (0.0061-0.0212) and -0.0032 (-0.0261-0.0201). A post-score check of partial correlations using the same Fisher outcome also left the eight-animal interval across zero. Slow outcomes had a larger residual scale. Thus the absolute coefficient contrast repeated, but a normalized gain did not clearly repeat. We provide original code, aggregate evidence, provenance hashes and a retrieval recipe. External records and record-level derivatives are not bundled. Maps and activity came from different animals. Both activity samples came from one lab. Unequal noise, shared input, sensor kinetics and atlas error remain possible explanations. The result does not establish a causal electrical filter or equal normalized effects.

C. elegans · gap junction · structure-function · calcium activity · temporal frequency · partial correlation · replication · outcome scale