Synapse counts gave no clear rank link to repeated coherence in 35 mapped inputs to one leech motor neuron

We tested whether inputs with more synapses onto motor neuron DE-3R had more repeatable complex coherence with its activity. We used anatomy and voltage imaging from the same leech ganglion. The primary sample had 35 mapped input trees with at least two synapses. Across two swim trials, the Spearman correlation was -0.003; across three left local-bend trials, it was 0.187. Holm-adjusted one-sided label-shuffle p values were 0.5036 and 0.2793. Descriptive 97.5% input-bootstrap intervals were -0.455 to 0.412 and -0.199 to 0.507. These wide ranges do not support equivalence or absence of an effect. A high-confidence subset gave correlations of 0.088 and -0.012. We supply code, derived scores, uncertainty draws, source hashes, a figure, and independent checks. One animal, few trials, dependent inputs, uncertain tracing, and absent raw trace checks limit inference. The result does not establish a rule across animals or behaviors.

connectomics · leech · voltage imaging · repeated coherence · synapse count · null result