Changes at other inhibitory puncta gave no clear gain in two-day size estimates under two rules in six mice

We tested whether size changes at other inhibitory puncta help to estimate the change of one surviving punctum. We used baseline hippocampal CA1 data from Klimmt, Kuhn, and colleagues. The eligible sample had 2,202 two-day site records from 520 episodes, 26 neuron-compartment groups, 17 neurons, and six mice. A fixed rule used the mean log-size change at the other survivors in the same compartment. A reference used other neurons in the same mouse, compartment, and interval. The fixed rule gave 2.9% more squared log-size error. A 95% mouse interval allowed 9.3% less to 15.0% more error. A later extension learned a partial correction from five mice and tested it in the sixth. Its mean error was 2.5% below the other-neuron reference and 1.0% below a no-change rule. These extension results are descriptive. Under a linear rule with area weights and independent unit-variance inputs, learned corrections gave 0.3% more response proxy error than the other-neuron reference. Currents were not measured. The test uses other sites from the second image; it is not a forecast. It excludes site loss and appearance. Common measurement effects and biological regulation cannot be separated. The results do not establish independent site changes. The package preserves source data, exact cell addresses, code, all predictions, figures, checks, and repeat instructions.

inhibitory synapses · synaptic size · gephyrin · hippocampus · longitudinal imaging · prediction · structure and function