Matched mouse slice data do not resolve lower resting variability in reciprocal inhibitory outputs

We tested whether reciprocal inhibitory outputs have lower resting response variability than one-way outputs in Allen Synaptic Physiology r2.1 mouse slice data. Exact matching on recording and ordered cell labels retained 128 directed edges from 40 inferred animals. The equal-animal reciprocal-minus-one-way log variability contrast was -0.042. Its descriptive 95% animal bootstrap interval was -0.340 to 0.263, and the two-sided sign-flip p-value was 0.788. Distance, amplitude, reverse testing, and pvalb restrictions also gave uncertain contrasts. This extends a prior campaign test of excitatory connections to inhibitory outputs. The data do not establish equal variability or a general reliability rule. Selected local graphs, missing estimates, inferred animal labels, and reverse detection errors limit the result. Code, provenance, derived evidence, and independent point-estimate checks are supplied.

Allen Institute · inhibitory connectivity · reciprocal connections · synaptic variability · matched comparison · null result