Ten annotated presynaptic sites join PD2 and PV5 olfactory target groups in one fly hemibrain at endpoint scores of at least 0.5

We tested whether olfactory outputs to cell groups with different measured response dynamics have separate presynaptic sites. We selected 33 projection neuron bodies and 17 target bodies from the fixed Kim et al. (2025) dataset. We then joined outgoing and incoming synapse annotations in the public hemibrain GCS v1.2 export. A site was an exact presynaptic coordinate within one source cell. Both endpoint scores had to be at least 0.5, and presynaptic x had to be less than 80 micrometres. The retained graph had 1,046 contacts at 1,016 sites and 16 target bodies. Ten sites contacted both PD2 and PV5 groups. DM1 and DM4 each supplied four sites; DP1m and VM3 each supplied one. A separate annotation-ID lookup confirmed all 20 contacts and their distinct postsynaptic points. Eight shared sites remained for LHPD2a1 and LHPV5a1_a alone. At a minimum endpoint score of 0.8, three broad-group sites and one narrower-class site remained. A repeat analysis of ten paired recordings from other flies gave a third response above the first in six PV5 recordings and below the first in four PD2 recordings. The anatomy gives counterexamples to complete separation of the selected output sites. It does not establish shared vesicle release or the dynamics of any individual site. This is one selected brain subset, and most retained sites contacted only one group. The package preserves the site inventory, response extracts, two figures, source generations and hashes, code, direct annotation records and checked repeat instructions.

Drosophila · olfaction · hemibrain · polyadic synapse · short-term plasticity · structure and function · data provenance