Missing strength estimates leave a majority strength-matching rule unresolved in 35 Allen mouse reciprocal excitatory pairs

We tested whether most detected reciprocal excitatory pairs have resting voltage responses within a factor of two. Allen Synaptic Physiology r2.1 small data supplied 35 eligible mouse visual-cortex pairs from 26 inferred mice. Only 21 pairs had two usable strength estimates; three met the rule. We kept incomplete pairs and allowed missing estimates to take any positive value. The sharp equal-mouse sample bounds were 8.97% to 52.88%. An approximate donor bootstrap outer range was 0% to 70.51%. Thirteen human temporal-cortex pairs gave sample bounds of 18.18% to 40.91%, with a wider sampling range. Thus the data do not resolve a majority matching rule in a larger population. The package includes code, all selected dyads, exact SQL, missing-data bounds, uncertainty draws, provenance, a figure, and an exhaustive bound check. Small slice samples, inferred donor groups, selection, and unmodeled response-fit error limit inference. This study tests a strength tolerance rule, not weight correlation, conductance symmetry, or intact-brain dynamics.

Allen Institute · reciprocal connectivity · synaptic strength · missing data · partial identification · null result