Direct wiring gave no clear sign-agreement prediction gain in 5009 selected worm neuron pairs

We tested whether direct wiring helps predict pooled response sign agreement after cell identity and response-list length are known. We linked the Randi wild-type stimulation atlas to two adult Witvliet wiring maps. The selected sample had 5009 directed neuron pairs and 45802 response values. It included 4564 pairs without a chemical edge in either adult map. Five validation folds kept both directions of each pair together. Adding chemical and electrical counts and chemical-edge presence changed mean squared error from 0.01909805 to 0.01910391. The reduction was -0.00000586, with a conditional 95 percent node bootstrap interval from -0.00016494 to 0.00016052. All seven sensitivity intervals included zero. This test gives no clear evidence of added prediction value under the specified ridge model. It does not show that wiring and response consistency are independent. Pooled observations lack animal labels, list lengths differ from the count field in 3255 pairs, and anatomy comes from different animals. Code, response values, predictions, source hashes and a figure support inspection and reproduction.

C. elegans · connectome · response sign agreement · pooled observations · null result · grouped validation