Reviewed all 51 source annotation rows, literature abstracts, retrieved full text for the critical interaction and trafficking decisions, the UniProt record, and the existing Falcon synthesis. Source IDs, evidence codes, and source row identities were preserved. Replacement terms GO:0038187 and GO:0004675 were checked against QuickGO definitions. Replaced research paraphrases presented as quotations with verbatim primary-source text. Electronic mappings use biological rationale and additional references; snippets are not required merely because the annotation is electronic.
The canonical function remains a plasma-membrane flagellin receptor kinase. Antibacterial defense is core, not a peripheral phenotype: PMID:18158241. The Rhea serine-kinase mapping is compatible with receptor serine/threonine kinase activity and should not be replaced merely because it captures only one substrate-residue class.
SCD1 interaction was initially inaccessible, but a pipeline refresh obtained full text and resolved it: PMID:20472560. This is association evidence, not proof of a direct binary interaction. No false SCD1-to-FLS2 attribution claim is warranted.
GRP7 physically associates with FLS2 protein in addition to binding its transcript; transcript regulation does not invalidate the protein interaction: PMID:23395902. FLS2 self-association is explicitly reported in the interaction network and is retained as non-core, rather than rejected because BAK1 forms a heteromeric complex: PMID:29320478.
Primary FLS2 trafficking evidence is distinct from BIK1 internalization: PMID:23085733. PMID:32404997 explicitly measures later FLS2 puncta as well as earlier BIK1 puncta; the annotations no longer rely on treating every BIK1 observation as an FLS2 observation. The callose and antibacterial defense annotations are directly supported in refreshed PMID:19095898 full text.
The CLV3p-FLS2 interpretation is disputed. PMID:21499263 reports noncanonical ligand signaling, whereas independent experiments report PMID:22923673. The associated IPI row is UNDECIDED rather than a confident noncanonical ligand claim; the original reference is marked DISPUTED. The replication paper was identified with Europe PMC and cached through the publication pipeline. This is not a retraction claim.
Validation passes with the advisory warning that annotations do not cite the deep-research report. That report was consulted as context; primary literature supplies the supporting text. There is no reason to add a synthetic report quotation to suppress this warning.
The required research refresh completed successfully with Falcon in 520.96 seconds (2026-09-11T20:48:53–20:57:34 UTC); fallback was not needed. Publication caching ran concurrently and cached all 28 original PMIDs. The new synthesis agrees with the canonical function and highlights 2024 nanodomain and synthetic-ligand studies; those report-level details were not promoted into additional GO annotations without primary-source adjudication. Publication-pipeline refreshes upgraded PMID:20472560 and PMID:19095898 to full text; PMID:22923673 was refreshed as abstract-only and newly added to this review.
Read reviewer comments 0 and 4 against the current annotation-reviewer policy. Changed all 19 GO:0005515 MARK_AS_OVER_ANNOTATED rows to REMOVE because generic binding does not describe a molecular activity; preserved each reported interaction and its verified quotation. The structural ligand-recognition row (PMID:24114786) remains MODIFY to pattern recognition receptor activity; the CLV3-dependent association row (PMID:21499263) remains UNDECIDED because the ligand interpretation is experimentally disputed (PMID:22923673). The reviewer's older request to use MARK_AS_OVER_ANNOTATED for generic binding is superseded by the canonical skill; neither term vagueness nor missing full text is used to declare an experimental interaction false. Association with BAK1/BIK1 or a regulatory/effector partner is not itself an assay of a distinct FLS2 molecular activity.
Reconsidered the earlier Rhea decision: serine phosphorylation remains chemically compatible, but MODIFY to GO:0004675 better records receptor signaling context, consistently with the other broad kinase rows and core function. This is an informative replacement rather than a claim of strict ontology descent or contradictory residue specificity. GO:0004675, GO:0038187 and GO:0010008 names/definitions were verified with QuickGO REST.
Restored endosome membrane in the receptor core with explicit trafficking scope [PMID:23085733, “FLS2 localizes to bona fide endosomes via two distinct endocytic trafficking routes depending on its activation status.”]. This does not claim a separately measured endosomal kinase reaction. Restored family/subfamily provenance using the actual cached PTHR48056-entries.csv Q9FL28 row and UniProt SF83 cross-reference. No ancestral-node assertion was guessed, and the InterPro metadata's unchecked LLM family description was not used as evidence. The primary paper and deterministic family membership serve different purposes.
All 51 original annotation identities were preserved. No research jobs, publication refreshes, or renders were run for this bounded follow-up.