With fixed trial order, 13 fly recordings showed different PD2–PV5 response contrasts between two odor orders after 20 s conditioning.

We tested whether a target-type cross-adaptation difference held in two odor orders. We used open fly recordings from Kim and colleagues. The main protocol had six PD2a1/b1 cells and seven PV5a1 cells, with 20 s conditioning at 5 Hz. We compared one-second conditioned and initial responses to the same test odor. The normalized response contrast was (conditioned minus initial) divided by their sum. The PD2-minus-PV5 contrast was -0.2618 for ethyl acetate then methyl acetate and -0.0014 in reverse. The planned interaction was -0.2604. Its descriptive 95% cell bootstrap interval was -0.4081 to -0.1031. An exact cell label test gave 23/1716, or p = 0.01340, under independent cells and exchangeable null labels. Date and trial resampling kept the interaction interval below zero. Small DM1 and DM4 input samples gave much smaller mean changes. Both input channels reached both target groups in a separate hemibrain source table. The trial audit found fixed recording order in all 29 matched trial-index pairs. Odor order and recording order cannot be separated. Animal identities were not established. These results constrain the tested response conditions, not a specific adaptation mechanism. The package preserves 186 trials, 33,670 spike times, source versions and hashes, new code, all results, three figures, and repeat instructions. A source-data rebuild and an independent exact count check passed.

connectome · Drosophila · olfaction · cross-adaptation · stimulus history · lateral horn · trial order · open data