## CDK5
- **UniProt:** Q00535 · **batch:** proteostasis-pr-1217 · **review status:** COMPLETE
- **PN placement:** 1 row, ALP — `Autophagosome closure maturation and lysosome fusion → "Specific function in autophagosome maturation and lysosome fusion unknown"`. **PN-node mapping:** leaf group=no_mapping ("unknown/residual"); class=context_only→GO:0016236; branch=no_mapping. **No GO propagates.** PN Notes: Drosophila Cdk5 phosphorylates ACN (acinus) at S437, stabilizing it and promoting basal autophagy.
- **Consistency:** Consistent and well-handled. PN places CDK5 in the residual "function unknown" autophagy-maturation bucket on the strength of the *Drosophila* Cdk5-Acinus basal-autophagy work (same eLife paper underpinning ACIN1's PN row). The review does not import that as a human fact; instead it anchors a human autophagy role on CDK5→SH3GLB1/endophilin B1 phosphorylation (GO:0016241 regulation of macroautophagy, ACCEPT, PMID:21499257). The notes (line 2860) explicitly state the Drosophila Cdk5-Acinus evidence was treated as context-only. No contradiction.
- **PN story / NEW pressure:** Already captured / PN over-reaches on the Acinus angle. PN's "unknown" autophagy-maturation placement is genuinely unknown for human CDK5; the review correctly mints no maturation term and routes the conserved-Acinus hypothesis into a suggested_question + suggested_experiment (does human CDK5 phosphorylate ACIN1 at the S437-equivalent?). The defensible human autophagy term GO:0016241 is already present. No ADD warranted.
- **Mapping strategy:** No change. The leaf's no_mapping is the correct conservative call — the node label literally says "function unknown," so propagating GO:0016236 macroautophagy would be TRAPP-like over-propagation. CDK5's inclusion does not justify upgrading the node; its real human autophagy role (SH3GLB1) is already individually annotated and is regulation, not maturation.
- **Evidence alignment:** Deliberately divergent. PN cites only the Drosophila eLife Cdk5-Acinus-S437 paper; the review's autophagy support is the human/neuron PMID:21499257 (endophilin B1), which PN does not cite. The divergence is the finding: PN's evidence is orthology-based, the review uses human evidence and quarantines the fly paper as a hypothesis.
- **Verdict:** Fully consistent; exemplary handling of an orthology-driven "unknown" placement (parallels ACIN1). No edits needed.