Activity peak epochs gave no clear extra count allocation after cell totals and overlap were fixed in one mouse PPC panel

We tested whether synapse allocation follows activity peak epochs beyond cell totals and anatomical overlap in one mouse posterior parietal cortex panel. We rebuilt peaks from saved calcium activity. The panel had 747 candidate directed pairs and 185 synapses among 127 cells. Activity epochs differed from saved information peak epochs in 82 of 134 valid source cells. This changed the same-versus-different class of 35 of 124 connected pairs. Most pairs still had different epochs. A conditional Poisson reference kept every source and target total fixed and used measured overlap. E-to-I same-epoch allocation was 20.69%, against 18.78% expected. I-to-E allocation was 23.72%, against 22.24% expected. Forward-minus-backward fractions were 17.24% and 12.18%, against 18.59% and 12.05% expected. All four primary Holm probabilities were 1. E-to-I tests used all 6,324 feasible tables. I-to-E tests used four chains and 80,000 saved tables. A right-trial check gave unadjusted p = 0.0098 among 62 sensitivity comparisons. It did not pass correction. A selected right-versus-left contrast gave p = 0.248. We also prove that mean signed lag cannot vary under fixed cell totals. Exact integer certificates verify the primary feasible limits. This study tests count allocation in one partial graph. It does not exclude a circuit role for inhibition or give uncertainty across animals. The package preserves code, derived inputs, source hashes, all reference scores, results, figures, plans, validation and repeat instructions.

connectome · structure and function · posterior parietal cortex · activity sequences · inhibition · conditional count model · data provenance