A 10 micrometre path test found no clear movie-response grouping of distinct inputs on 237 dendritic trees from one MICrONS mouse

We tested local input grouping in one mouse visual cortex map. We joined measured mean movie responses to a preserved MICrONS path-distance dataset. Exact nucleus and optical recording matches retained 3,131 of 3,437 selected sources. The primary test used 237 target cells, 718 sources, 529 near pairs, and 1,269 reference pairs. Near paths were at most 10 micrometres. Reference paths were greater than 20 and at most 50 micrometres. Each distinct source pair entered once per target. Targets had equal weight. Source response labels were exchanged within area, cell type, and 100 micrometre soma cubes across the full graph. The excess near-input Pearson correlation was 0.00972. Its upper-tail probability was 0.1946 from 4,999 draws. The central 95% target resample range was -0.01291 to 0.03268. A planned reviewed-axon check gave excess 0.05393 and probability 0.017 on 58 targets. Exploratory finer spatial and optical-scan controls reduced that excess and weakened its evidence. Including contacts from one source raised the raw contact-pair contrast from 0.0110 to 0.1677. Those contacts share one response label by construction. Independent formulas, official pair values, source checks, and a clean repeat passed. All 72 numerical output files matched exactly. The package preserves code, measured response extracts, anatomical distances, provenance, all results, figures, and repeat steps. The study has one mouse, selected sources, sparse input coverage, finite response reliability, and approximate contact attachment. It does not measure spine responses or establish absence of functional input groups.

MICrONS · dendritic input · functional connectomics · measured movie responses · signal correlation · source identity · skeleton path distance · spatial controls · recording controls