Occurrence matching left the wiring-related strain contrast unresolved in 504 worm cell pairs

We tested whether direct chemical wiring predicts a different change in response detection between wild-type and unc-31 worms. We linked the Randi signal propagation atlas to two adult Witvliet maps by cell name. Exact matching of the reported occurrence count retained 504 of 5,151 common tested directions. Of these, 43 had a chemical edge in either map. The wiring contrast in mutant-minus-wild-type detection was +0.43 percentage points, with a nominal 95 percent node bootstrap interval from -3.61 to +4.58. Only two wired pairs had detection; both did so in both strains. The empirical bootstrap cannot represent unseen wired transitions. A count audit found that 114 of the 504 occurrence-matched pairs had unequal stored response-array lengths. Stored-length matching also left the contrast unresolved. These data do not establish a wiring-related strain effect or its absence. Count matching selects a small sample and does not ensure equal detection power. Different anatomical and functional animals, aggregate q values, few events and conditional resampling limit inference. Code, paired evidence, source hashes, retrieval steps and a figure support reproduction.

C. elegans · structure-function · unc-31 · measurement coverage · count audit · null result