## HSPB9
- **UniProt:** Q9BQS6 · **batch:** proteostasis-batch-2026-06-07b · **review status:** COMPLETE
- **PN placement:** one row — `Cytonuclear proteostasis|Chaperone|small HSP system|small HSP` (type) ; **PN-node mapping:** sHSP type → mapped/ok_for_propagation_to_go GO:0044183 protein folding chaperone (new_to_goa); group/class/branch no_mapping.
- **Consistency:** Notes ↔ review consistent — HSPB9/CT51 is a poorly characterized, rapidly evolving, testis-restricted sHSP; its only experimental molecular interaction is with the dynein light chain TCTEL1/DYNLT1 (PMID:15503857, "potentially interacts"), and direct chaperone/holdase activity has NEVER been demonstrated for this paralog (PMID:11470154). The review explicitly declines to assert a confident holdase MF, making GO:0045503 dynein light chain binding the core MF and flagging chaperone activity as inferred-only.
- **PN story / NEW pressure:** GOA has no chaperone MF (only GO:0005515 protein binding). PN projects GO:0044183 protein folding chaperone (verified real, foldase per OLS) purely by family membership. This double-over-reaches: (1) foldase vs holdase term mismatch shared with the other sHSPs, AND (2) HSPB9 has no demonstrated chaperone activity of any kind, so even a holdase term (GO:0140309 / GO:0051082) would be inference-only, not evidence-backed. The review's defensible MF is GO:0045503 dynein light chain binding (verified real, from PMID:15503857).
- **Mapping strategy:** HSPB9 is the weakest member for the sHSP-family GO:0044183 mapping — both the wrong activity class (foldase) and unproven for this paralog. If the family node propagates any MF, it should be a holdase term carried as inference (IBA-grade), not a foldase; better to treat HSPB9 as an under-characterized exception.
- **Evidence alignment:** PN row has no reference titles; review PMIDs (11470154 identity/testis; 15503857 DYNLT1 interaction/CT antigen; 19464326 localization) — no overlap to diverge, but none supports folding-chaperone activity, undercutting the PN projection.
- **Verdict:** sHSP placement defensible by homology but the GO:0044183 foldase projection over-reaches (wrong activity class + no demonstrated chaperone activity). **Recommended edits:** [MAP] do not project GO:0044183 (foldase) onto Q9BQS6; if the sHSP node propagates an MF use GO:0140309 / GO:0051082 as inference, and note HSPB9's only experimental MF is GO:0045503 dynein light chain binding.