Digital Drosophila Archive
2.0 KiB
Phase 4X — evidence and interpretation limits
Checked 2026-09-29. No external IDs replace MaleCNS canonical IDs.
- Author repository funkelab/synister_malecns describes MaleCNS transmitter prediction and links its ground-truth source. A prediction pipeline is not a measurement of every cell's transmitter or receptor-dependent current sign.
- flyconnectome/drosophila_neurotransmitters is the linked author ground-truth resource. This audit does not import updated labels or infer correspondence from names alone.
- Haynes et al., eLife 2015, doi:10.7554/eLife.03868, reports DPM GABA/5-HT signaling and inhibitory effects in its tested preparation. This motivates a DPM interpretation warning; it does not assign a replacement sign to these exact MaleCNS bodies.
- Eckstein et al., Cell 2024, doi:10.1016/j.cell.2024.03.016, discusses DPM prediction discrepancies in its studied datasets. Prior 4W source-access limitations remain recorded; no claim of having retrieved the blocked full text anew.
predicted_nt_confidence and celltype_predicted_nt_confidence retain their actual raw names and numeric values. No independently calibrated probability interpretation, confidence threshold, or physiology-based sign correction is introduced.
Raw ground_truth is reported as a supplied field. Agreement of consensus with that field is not an independent test of prediction accuracy. Exact annotation class DAN, exact type DPM, transmitter metadata, and prior exact-compartment connectivity mapping are distinct evidence axes. None alone establishes appetitive/aversive function, dopamine concentration, or the current-sign response of every target.
This fixes audit semantics, not biology. Frozen runtime semantics, including dopamine=+1, remain unchanged. Any alternative receptor/modulatory model requires a separate explicit hypothesis and validation.