F13B

Summary: Retain experimentally supported secretion of the B subunit.

Reason: PMID:26247044 measures wild-type and mutant human FXIIIB in cleared extracellular medium after HEK293T expression (Methods: cell culture and protein quantification; Results and Figure 1). Wild-type secreted B also reconstitutes A2B2 with added recombinant A2. This directly supports extracellular region. Intracellular ER/Golgi images concern biosynthetic transit and mutation-associated retention, not a replacement for the secreted wild-type location. The expression preparation does not establish function of the separate predicted RefSeq products.

Summary: Retain the established extracellular plasma pool.

Reason: The original PMID:4405643 is the verified human plasma/platelet factor-XIII study, but its complete Methods and Results were not recovered here. Retain the existing positive experimental location because independent human B-chain secretion and the exact P05160 plasma-proteome entry support it (PMID:26247044; PMID:14718574 Table II). This is independent corroboration, not a claim to have reread the 1973 purification experiments or to localize B to platelets.

Summary: Retain the verified secreted-location mapping.

Reason: The actual current P05160 record assigns Secreted (SL-0243) with PMID:26247044 and PMID:4405643 evidence. The official SL-0243 entry maps directly to GO:0005576. Human B-chain secretion into cleared culture medium provides independent experimental support; no membrane anchoring or alternative-product localization is inferred.

Summary: Retain the target-specific plasma-proteome location.

Reason: The recovered original PMID:14718574 Table II explicitly lists P05160 as factor XIII B chain, with presence in the literature, 2DEMS and LCMS1 inputs and absence from the LCMS2 input. The inspected Data Sources and Methods identify human serum/plasma proteomic sources. This supports the extracellular pool, together with directly measured B secretion. Only selected original text and the target table were available; no unique peptide, intact mature-chain sequence or predicted-product assignment is claimed.

Summary: Retain the native B-subunit extracellular location in the FXIII activation model.

Reason: The actual Reactome R-HSA-140599 input 140608 and cleaved output 140789 both contain human F13B entity 140605, whose own compartment is extracellular region and whose reference is P05160, residues 21–661. Thrombin is the proteolytic catalyst and the A chains are cleaved; F13B is an associated component. Thus this location is read from the target entity, not copied from the reaction compartment or mistaken for B-chain protease activity.

Summary: Retain extracellular localization of complexed and released F13B.

Reason: Reactome R-HSA-140847 contains F13B 140605 within the cleaved A2B2 input and releases two copies of that same P05160 21–661 entity, explicitly in extracellular region. The activated output 140849 contains A chains and calcium, without B. Retain the B location while distinguishing its release from A-chain activation and catalysis.

Summary: Retain the phylogenetically supported coagulation role.

Reason: Human P05160 lies at PTN002499773 below the actual positive GO:0007596 IBD at PTN001455694 in PTHR45785; its aligned sequence is exactly the displayed 661-residue sequence and no target-path IBD NOT contradicts this process. The IBD cites human P05160 and mouse MGI:88379: self-WITH is valid descendant experimental grounding. Human B secretion and A2B2 assembly in PMID:26247044 support a carrier/assembly contribution, while the positive mouse knockout and human-B rescue abstract of PMID:18224415 supports the conserved role with full Methods unavailable. This does not assign A-chain cross-linking chemistry to B. Blood coagulation remains ACCEPT as a core biological role: the core records its supported fibrin-clot-formation part rather than repeating both terms; omission of the broader term from that list does not make coagulation peripheral.

Summary: Retain the ortholog-supported noncatalytic coagulation role.

Reason: The current Q07968 record identifies mouse F13b and the literal ENSMUSP00000027615 donor. Its experimental blood-coagulation IMP traces to PMID:18224415, whose abstract reports reduced plasma cross-linking in B-deficient mice and rescue after recombinant human B administration. The complete donor Methods were not available. Independent human B reconstitution of A2B2 in PMID:26247044 and the human deficiency/stability observation in PMID:2334637 support the target role without asserting an intrinsic transglutaminase activity or transferring the propagated reverse human-to-mouse assertions as experiments. Blood coagulation remains ACCEPT as a core biological role: the core records its supported fibrin-clot-formation part rather than repeating both terms; omission of the broader term from that list does not make coagulation peripheral.

Summary: Retain the human coagulation carrier/stability context.

Reason: PMID:2334637 is the verified familial B-subunit deficiency report. Its accessible abstract describes absent B, reduced plasma A and a short disappearance half-life of infused A, while platelet A was normal; full original Methods remain unavailable. This positive clinical evidence is consistent with the independently measured human B–A2 assembly in PMID:26247044. Retain the existing TAS process as a noncatalytic support role in coagulation, not evidence that B itself makes fibrin cross-links or that a deficiency phenotype alone identifies the catalytic performer. Blood coagulation remains ACCEPT as a core biological role: the core records its supported fibrin-clot-formation part rather than repeating both terms; omission of the broader term from that list does not make coagulation peripheral.

Summary: Retain clot-formation participation, with a bounded query about the source assay attribution.

Reason: F13B supports fibrin-clot formation through its noncatalytic carrier/assembly role. Human B reconstitutes A2B2 with recombinant A2 (PMID:26247044); the original abstract of PMID:18224415 additionally reports recovery of plasma A and cross-linking after human B administration to B-deficient mice. In the inspected PMID:27363989 article and supplement, human A-chain variants expressed in COS-1 cells are the experimental targets and commercial A2B2 is a positive control. That whole-complex evidence can legitimately support component-level process participation, without assigning transglutaminase chemistry to B. No demonstrated evidence-code defect is established by the absence of an isolated B perturbation; the imported IDA is retained with this complex-level scope.

Summary: Retain membership of the plasma factor-XIII enzyme system.

Reason: The actual PMID:27363989 Introduction describes A2B2 and its activation, and the supplementary alpha-2-antiplasmin assay uses A2B2 as the positive reference preparation. Independently, PMID:26247044 directly reconstitutes secreted B with recombinant A2 and measures heterotetramer antigen. Retain transferase-complex membership for this proenzyme assembly; the complex-level term does not make every subunit a transferase. Reactome distinguishes the B-containing zymogen from the activated A-plus-calcium output after B dissociation, so this annotation is not a claim that B persists in the catalytic product.

Summary: F13B contributes a noncatalytic structural role to the circulating A2B2 precursor complex.

Reason: The current GO:0005198 definition concerns contribution to the structural integrity of a complex. In the inspected human B-subunit experiment, wild-type and mutant constructs were expressed in HEK293T cells, a fixed amount of recombinant A2 was added to secreted medium, and the resulting A2B2 antigen was normalized to total B antigen. Multiple B variants reduced heterotetramer reconstitution relative to wild type, whereas Leu116Phe and Val217Ile retained comparable assembly. This B-dependent assembly phenotype supports the broad structural contribution of B to the precursor complex; IMP reflects the mutant comparison rather than asserting an isolated direct measurement of a molecular stabilization mechanism. This is not a generic inference from binding alone. The assembly endpoint does not establish kinetic resistance to dissociation, protection against degradation or aggregation, or B catalysis. B is released during activation, so the term applies to the B-containing precursor and does not assign a contributes_to transglutaminase activity in the active A complex. Same-role plasma carriers do not establish a universal annotation convention: current exact/descendant queries found no structural term on IGFALS or IGFBP3; VWF has an extracellular-matrix structural child term for a distinct role. Those observations neither prove this assignment nor prohibit the target-specific experimental inference. No existing molecular-function assertion is an ancestor or descendant of the proposed term.