Electrical coupling calls showed extra dependence in 91 adult mouse TRN maps under a fixed distance and recording state model

We tested whether electrical coupling calls from one recorded neuron were independent after measured sampling controls. We reused a signed reconstruction of public adult mouse thalamic reticular nucleus data. The main set had 415 neighbor tests, 91 recorded neurons and 78 calls. A fixed logistic null kept soma distance, source label, light correction method, baseline spike latency and baseline variation. The sum of residual products within maps was 10.768. Thirty-six of 9,999 null refits reached that value. The upper-tail p-value was 0.0037. All ten planned checks gave p-values at most 0.014. Removing separate blank subtraction gave p = 0.010. Deleting the map with the largest contribution gave p = 0.013. A normal map effect gave a median odds ratio of 2.41, with an approximate 95% profile interval of 1.37 to 4.36. Yet 43 maps without calls were close to 42.6 expected. The empty-map total alone missed extra dependence. Shared recording state, date, animal or assay error can cause dependence. These data do not isolate a stable neuron trait or anatomical gap junction degree. The package supplies code, trial tables, provenance, null draws, figures and clean offline repeat steps.

electrical synapses · thalamic reticular nucleus · spike timing · functional connectivity · sampling · residual dependence · parametric bootstrap