Binary input patterns had closer nearest neighbours than degree and broad cell type controls in eight partial worm maps

We tested binary chemical input patterns on a fixed panel of 162 neurons in eight public C. elegans brain maps. The distance between two target patterns counts sources that connect to only one target. Mean nearest distance was 7.1 to 12.2 percent below degree-control means. No primary control draw reached the observed value in 1,000 draws per map. Longer degree-control runs gave the same order. An exploratory control that also kept each neuron's input and output counts by broad cell type gave the same order. Exclusion of left-right partner comparisons also kept this order. We prove that nearest distance is the minimum number of known source erasures that can make a target input pattern equal to another one. In contrast, mean all-pair distance is fixed by source degrees. These are structural model results, not measured coding or activity results. Spatial contact, partial coverage, binary edges, and different animals at each age limit interpretation. Code, graph matrices, source hashes, all control scores, graph snapshots and checks are included.

connectomics · C. elegans · input patterns · source erasure · degree controls · Hamming distance