Contacts to different cells at 52 shared sites in one fly larva showed no clear extra area similarity after connection controls

We tested whether contacts to different target cells at one presynaptic site have similar areas after controls for each connection. We used 540 measured contact areas from 80 connections in one first-instar female fly larva. The main test used 72 contact pairs at 52 shared sites from 18 source cells. We averaged squared log-area differences within sites, then within source cells, then equally across source cells. We shuffled areas within each directed source–target connection. This kept all connection area distributions and each site's contact and target composition fixed. The observed score was 0.8274. The exact reference mean was 0.8193. A test with 19,999 shuffles gave two-sided p = 0.9003. Depth control, removal of connection means, and checks of repeated contacts gave no clear evidence for extra area similarity. The median larger-to-smaller area ratio was 1.92. Under a separate linear model with area weights, equal contact weights gave a median error of 15.8 percentage points in pair-normalized target shares. This is a model result, not measured transmission. The data concern selected connections in one animal. Wide source resample ranges do not exclude moderate size similarity. We supply a paired area benchmark, exact reference means, all shuffle results, figures, source data and hashes, and repeat instructions.

connectome · polyadic synapse · synaptic area · Drosophila · structure and function · conditional randomization