With equal leak and free cell gains, 25 Cook neuronal gap models failed to fit 218 C. elegans calcium means within 2.61 sample standard errors

We asked whether a passive electrical wiring model can fit positive calcium means when each cell has a free input gain and readout gain. We used 218 directed means from 109 selected C. elegans response pairs. A repeated free reciprocal fit allowed changes within 1.2410 sample standard errors. Four native gap exports had missing routes that set larger fit limits. We therefore added an exploratory test of the corrected July 2020 Cook matrix. We retained all 302 neuronal slots and all available neuronal gap weights. The static model used equal leak and 25 fixed gap scales from 0.001 to 1000. With recorded weights, its best scale needed a common band of 2.6174 sample standard errors. All 25 scales failed at 2.61. Separate graph and linear-program checks agreed, and we preserve a failed-cycle certificate for each scale. One-per-link and square-root weights needed 2.6535 and 2.6726. Retaining all 469 named cells in the worksheet gave 2.6706 with recorded weights. A matched exploratory sample of 118 means had routes in all five maps; its best recorded-weight bands ranged from 1.5287 to 2.2290, against a reciprocal band of 1.2410. These are deterministic sensitivity sets, not confidence intervals across animals. The export lacks animal and event identifiers. The result limits the stated static models and does not identify a biological cause. The package preserves original code, selected response facts, complete retained graphs, fitted means, failure certificates, figures, source hashes and repeat instructions.

connectome · C. elegans · electrical synapse · passive network · cell gain · signal propagation · difference constraints · sensitivity analysis