Uniform samples of 588 synapse sites gave 668 and 678 expected connections in two adult C. elegans graphs of 162 embryonic neurons

We tested chemical connection counts at equal synapse site counts in eight published C. elegans brain graphs. The main analysis kept 162 embryonic neurons and removed self-connections. We sampled whole sites uniformly without replacement. We kept all retained partners at each site. The birth graph had 588 sites and 675 directed connections. The two adult graphs had 3,168 and 3,253 sites and 1,597 and 1,632 connections. At 588 sites, their exact expected counts were 667.67 and 678.34. These are 1.09 percent below and 0.49 percent above the birth count. Across all common sample sizes, adult differences ranged from -4.95 to -1.09 percent and -1.71 to 0.49 percent. Adding neurons born after hatching gave adult increases of 8.20 and 8.93 percent at 588 sites. About half of the expected adult pairs were not recorded at birth. The package supplies checked code, source hashes, results, and figures. Exhaustive tests, independent calculations, and fresh source runs passed. This exploratory result applies to artificial samples of fixed observed graphs. It does not show that connection growth is a sampling error or that wiring rules are unchanged.

C. elegans · connectome · development · synapse sites · polyadic synapses · rarefaction · sampling · reproducible research