Shared synapse sites increased target count variance in eight C. elegans maps under an independent site loss model with probability 0.5

One chemical synapse site can contact several target cells. We tested how this grouping changes the number of targets that retain direct input from one source cell. We used the eight brain maps from Witvliet and colleagues. Each site had an independent loss state. At an assumed loss probability of 0.5 per site, recorded groups gave 1.54 to 2.02 times the target count variance of independent contacts. Both models gave the same expected count. Recorded groups also exceeded the central 95 percent range of matched random groups in all eight maps. The random groups kept every connection count and every site target count fixed. Four finite chains gave 512 reference draws per map. These ranges are model references, not biological confidence intervals. All data selection checks kept the variance ratio above one. A model with partial sharing gave a smaller effect. Two exact examples showed that identical connection counts and site target counts can give different variances. Separate matrix calculations checked 211,004 pair records and 40,494 source records. The package contains code, sufficient counts, source URLs and hashes, reference draws, figures, and rebuild instructions. The study is exploratory. It does not estimate release rates or neural activity. The result applies to direct input availability under the stated loss assumptions.

connectome · C. elegans · polyadic synapses · synapse loss · target count · joint failure · reproducibility