Cell body distance gave no clear gain for rise time prediction in ridge tests of 1,572 mouse cortical connections with separate donors

We tested whether cell body distance improves prediction of synaptic response rise time in local mouse cortical recording graphs. We used 1,572 directed chemical connections from 570 inferred donor groups in Allen Synaptic Physiology r2.1. Each connection had voltage and current rise measurements. Five outer test folds and three inner folds kept each donor together. Baseline ridge models used cell labels and recording controls. Adding linear distance raised squared log voltage-rise test error by 0.12%. A central 95% paired donor range ran from 0.34% higher error to 0.11% lower error. This range holds test predictions fixed and omits refitting uncertainty. The matched current-rise and ratio ranges also included zero gain. A later control on 1,441 pairs gave 3.33% lower onset-delay error and 0.14% higher voltage-rise error. A later comparison that added measured current rise to the baseline gave 12.12% lower voltage-rise error. The tested distance rule gave no clear rise time benefit. The result does not establish that synapse location has no effect. Cell body distance is not dendritic path length. The package has code, model states, results, figures, source provenance and repeat instructions. Source measurements remain external under Allen terms.

functional connectivity · synaptic kinetics · rise time · cell body distance · mouse cortex · paired recordings · prediction · donor validation · negative result