With neuron totals fixed, an early wiring model gave a higher forward synapse fraction than each of two adult C. elegans maps

We tested whether the rise in the forward chemical synapse fraction across eight C. elegans brain maps exceeds a model with early pair affinities and each animal's neuron input and output totals. The main panel had 162 embryonic brain neurons. We used the class order sensory, modulatory, interneuron, motor. Three early maps selected a uniform mixture of 0.1 by omitted-map validation. The mean recorded forward fraction rose from 49.07% in the early maps to 55.39% in two adults. The fitted adult mean was 56.64%. The recorded adult mean was 1.26 percentage points below the model. A central 95% range from 1,000 complete-site resamples was -1.88 to -0.67 points. This range concerns the same maps, not a population of worms. Smoothing, early reference, neuron panel, label, and birth contact checks kept a negative residual. Flow bounds showed that the totals did not force the forward fraction. We also give the identity that fixes mean rank displacement at fixed totals. Adult neuron totals are inputs to the fit, so this is not a forecast from early wiring alone. The result limits inference from the global forward fraction; it does not exclude changes at specific connections. The package preserves code, source hashes, fitted matrices, tables, figures, resample results, integer optimization certificates, and repeat steps.

connectome development · Caenorhabditis elegans · feedforward wiring · neuron totals · iterative proportional fitting · synapse counts