With five observed targets, broad anatomy controls left no clear added electrical graph gain in a fixed calcium completion test for 83 C. elegans worms

Can electrical wiring help estimate calcium responses when most target cells are hidden? We used a fixed stimulation data table from Randi and colleagues. The test had 13,017 target responses from 506 stimuli in 83 C. elegans worms. Each random panel kept five target cells observed and hid every other target response at once. Five outer animal groups and three inner animal groups kept test worms out of fitting and parameter selection. A common response term reduced mean squared error by 3.13% from a baseline with cell identity, recording controls and direct wiring. In the planned test, an electrical graph covariance term gave a further 0.99% decrease beyond common activity and one soma position kernel. Its conditional 95% animal range was 0.13% to 2.02%. Later controls used more position scales, bilateral cell classes, pharyngeal membership and cell connection weight. The added graph gain then fell to 0.16%, with a range from -0.20% to 0.51%. The second mask seed and a shorter response window gave positive gains under the original controls. The red-only outcome gave no clear graph gain. These results support sparse response completion in this selected recording panel. They do not show a clear benefit from detailed electrical topology beyond the broader controls. The outcome is a fluorescence summary, and the anatomy came from other animals. The package preserves the input facts, disjoint masks, fitted states, predictions, full parameter grids, source hashes, code and repeat instructions.

connectome · C. elegans · gap junction · calcium imaging · missing response · animal holdout · graph kernel · structure and function