Shared site failure reduced two-step route survival in eight worm maps under equal contact survival

We tested random route survival in eight public C. elegans brain maps. We kept 141 cells with an output annotation in all maps and tested all 19,740 ordered cell pairs. One model failed each annotated chemical site as a unit. A reference failed each contact independently with the same survival probability. A proof shows that shared sites cannot increase direct or two-step route survival under these assumptions. At contact survival 0.2, the mean reduction relative to the reference was 1.38 to 4.30 percent across maps. Exact reference probabilities and conditional Monte Carlo estimates gave the map results. Small circuit enumeration checked the method. The ordering is a property of the model, not evidence of special biological wiring. Unknown release rates, ignored electrical links, partial maps, and one animal at most ages limit use. These are structural model results, not activity measurements.

connectome · C. elegans · polyadic synapses · route survival · shared failure · conditional Monte Carlo