At fixed neuron totals, early directions cut adult count deviance by 2.4–4.6% in two C. elegans maps with matched contact smoothing

We tested whether early chemical synapse directions add adult count information beyond contact area and neuron totals. Three early C. elegans maps supplied the templates. Two adult maps supplied evaluation counts and every neuron input and output total. A symmetric template kept each early unordered pair count but removed its direction. Both templates mixed early contact area and synapse counts. Settings came only from early leave-one-map-out tests. The selected directed family reduced adult Poisson deviance by 6.35% and 6.74% against the selected symmetric family. With the same settings in both templates, reductions were 2.43–4.57%. At shared directed settings, direction allocation supplied 94.0–99.5% of the total gain. Pure count templates and a birth-only mixed reference did not retain the gain at their tested settings. Stronger supplied counts by partner role reduced the total gain. One common setting then gave almost no gain in one adult. The direction score still improved. These results identify conditional direction information in pooled early maps. They do not establish a birth-only forecast, a causal wiring rule, or a population effect. Independent extraction, a second optimizer, exact score identities, fresh source hashes, and a clean repeat checked the results. The package preserves code, numerical facts, model matrices, figures, provenance, and repeat steps.

connectome · Caenorhabditis elegans · development · synapse direction · membrane contact · count allocation · iterative proportional fitting