Bacterial type V secretion

A species-neutral module for delivery of a passenger protein to the surface of a diderm bacterium by the type V route. The precursor crosses the inner membrane through Sec and is held in a translocation-competent, unfolded state in the periplasm by holdase chaperones. A beta-barrel translocator, which is either a domain of the passenger's own polypeptide or a separate partner protein, is installed in the outer membrane by the BAM complex. The passenger then crosses the outer membrane through that barrel and folds on the far side. No ATP or ion gradient is available at the outer membrane, so vectorial folding of the emerging passenger supplies the directionality. Sec translocation itself, the biological activity of the displayed passenger, and the general biogenesis of outer-membrane proteins unrelated to secretion lie outside this module.

## Boundary The module begins with a Sec-delivered precursor in the periplasm and ends with the passenger folded on the outer face of the outer membrane. Three things are deliberately excluded. Sec translocation is a shared upstream route, as in the sibling type II module. The passenger's own biological activity is excluded because the same architecture carries proteases, adhesins, lipases, and toxin-delivery proteins, so admitting passenger chemistry would make the module a list of unrelated functions rather than a description of a secretion mechanism. General outer-membrane protein biogenesis is excluded even though BAM is shared with it: BAM is admitted only for the specific step of installing the type V translocator. The periplasmic holdases are included rather than treated as upstream because the precursor's unfolded state is a requirement of the transport step itself, not merely of its delivery. They are nonetheless not type V specific, and the module asserts no exclusivity for them. Ontology and curation gaps are recorded in these notes rather than as `knowledge_gaps`, which CLAUDE.md reserves for biological unknowns that only new experiments could close. None of the items below is a biological unknown: a term request or a curator would resolve each one. ## Ontology gap: GO:0015474 hard-codes the classical domain order GO:0015474 autotransporter activity hard-codes the classical domain order, defining the passenger as N-terminal and the porin as C-terminal. The inverted architecture is genuine autotransport by mechanism but is excluded by that wording, so it has to borrow the generic transmembrane protein transporter term and the self-transport claim is lost. **Asserted instead.** Apply GO:0015474 only to the classical, trimeric, and patatin-like variants; use the generic transporter term for the inverted variant and record the reason on the annoton rather than stretching the definition. ## Curation and ontology gap: neither insertase carries a specific molecular function in GOA For both BamA and TamA the role is expressed only through process and component terms, leaving protein binding as the molecular function annotation. For BamA the gap is now argued to be a propagation failure rather than a curatorial judgment, since the equivalent mitochondrial insertase SAM50 carries the term by mutant phenotype and was never placed as an ancestral node from which it could descend to the bacterial clade. For TamA no such argument has been made. **Asserted instead.** Keep membrane insertase activity on both annotons and keep the note that GOA does not yet carry it. For BamA cite the focused gene review, which proposes the term with a comparator check; for TamA leave it as curator synthesis, which is all the evidence currently supports. The insertase molecular function is settled on the gene instead: `genes/ECOLI/bamA/bamA-ai-review.yaml` proposes GO:0032977 as a NEW annotation with a comparator check, and records the RcsF case it cannot cover as a term request.