Direct counts gave no clear gain for early minus late calcium responses in a ridge test of 104 C. elegans records

We tested whether direct chemical and electrical connection counts add information about early versus late calcium responses after neuron stimulation. We checked all 110 processed files in DANDI 001075 version 0.240930.1859 against their source hashes. Fixed rules retained 23,393 trial windows and 19,811 ordered pair means from 104 animals. The early window was 1 to 6 s. The late window was 15 to 20 s. The outcome was the difference between transformed relative green/red changes. Five outer folds excluded whole animals. Both ridge models had source and target identity terms and baseline controls. The full model added log direct counts from two adult connectomes. Mean squared error fell from 0.01503607 to 0.01502215, a 0.093% reduction. The descriptive 95% animal bootstrap interval was -0.121% to +0.295%, with predictions fixed. Stated label, distance, map, penalty, and direction checks also gave no clear gain. An added exploratory source-activation screen retained 55 animals and gave a 0.288% error increase. These results apply to the stated models, windows, and selected records. They do not show that wiring lacks timing information. The package preserves derived data, predictions, code, source hashes, independent checks, figures, and repeat instructions.

C. elegans · connectome · calcium imaging · signal propagation · response timing · animal exclusion · open data