Target pair counts at birth had excess agreement with two adult C. elegans brain maps under local count controls

We tested whether synapse target groups agree between one birth map and two adult C. elegans brain maps. A group has one source and two distinct targets at the same chemical site. Random adult maps kept all source-to-target counts and each site's target count fixed. The local model also fixed target counts within each source and 2 micrometre cube. Recorded birth-to-adult count overlap was 15.2 and 19.6 percent above the local random means. After each map's expected group counts were removed, agreement scores were 101.8 and 143.1. No random score reached them in 2,000 adult draws each. Removing both RIA sources left positive scores of 25.0 and 38.0. Grid shifts, merged source nodes and longer chains kept the count result. Distinct shared group counts gave a less stable result: one adult was inside the random range after a grid shift. We reproduce an earlier adult-to-adult SCI result and extend the comparison to birth. The focus changed after the first results when a later search found that adult study. This investigation is exploratory. Fine contact geometry and annotation can explain residual agreement. Three cross-sectional maps do not show individual synapse survival, shared release or a population trend. The package has code, source URLs and hashes, tables, random score draws, figures, exact sampler checks, audits and repeat instructions. A fresh-input replay reproduced the primary numerical outputs exactly.

connectome · Caenorhabditis elegans · development · polyadic synapse · target groups · fixed-margin randomization · higher-order connectivity