Matched Allen slice circuits do not resolve a source-versus-target difference in resting voltage magnitude

We compared resting synaptic voltage magnitudes among connections that share a source neuron or a target neuron. We matched recording, synapse type, cell class, Cre label, and layer in Allen Synaptic Physiology r2.1 data. The primary test retained 188 mouse visual-cortex connections from 40 inferred animals and 40 human temporal-cortex connections from seven inferred donors. Source-to-target squared log-difference ratios were 0.876 and 2.422. The 95% donor intervals were 0.728 to 1.753 and 0.987 to 6.098. Neither edge-value permutation test passed correction for two comparisons. Excluding one mouse response below 10 microvolts changed the mouse ratio to 1.225. The human excitatory subset gave a nominal signal, but it was a secondary comparison in six donors. These data do not establish equal source and target roles or a common rule across samples. Small selected graphs, uncertain response fits, inferred donor labels, and different circuit mixtures limit inference. The package includes code, selected edge records, resampling results, a figure, and independent score checks.

Allen Institute · synaptic strength · shared neurons · matched circuits · null result