Shared site loss lowered bilateral joint reachability at failure rate 0.5 in eight partial worm map models

We tested whether shared chemical synaptic site loss can improve joint route survival to bilateral targets in eight public C. elegans brain maps. We used 162 fixed neurons, 75 named bilateral pairs and 12,000 source-pair events per graph. Independent site loss and independent contact loss had equal marginal edge survival. There were 192,000 sampled graphs over three assumed failure rates. At the primary rate of 0.5, site loss lowered mean joint reachability by 0.31 to 3.82 percentage points. All eight approximate simultaneous Monte Carlo intervals excluded zero. It also raised exactly-one-target success in the two adult maps. Exact toy cases showed that joint differences can have either sign. This is a finite structural model test, not a measurement of release, activity or behavior. Partial maps, assumed failures, missing signs and electrical links, and sparse animal sampling limit transfer. New code, aggregate draw scores, source hashes, a figure and retrieval steps support reproduction. External raw records remain outside the bundle.

connectomics · C. elegans · polyadic synapses · joint reachability · correlated failures · bilateral neurons