This review preserves all 83 seeded source assertions, their partners and qualifiers. The source TSV has 84 rows; the two date-separated records for the same mouse-supported synapse assertion collapse to one seeded review object. The normal source log's API count of 85 is recorded separately from these actual output counts. No alternative-products slot was supplied, and none is invented. UniProt P02745 is the reviewed human 245-residue C1QA precursor; its secondary accessions are B2R4X2 and Q5T963.
The ordinary Falcon deep-research command, with the prescribed perplexity-lite fallback and installed client 0.2.7rc1, attempted both providers once. Both failed at DNS resolution and produced no provider file. This journal records manual reading; it is not represented as provider-generated research. Normal source caches and existing cache variants are immutable. The two previously missing historical sources were imported through the C1QB normal-source route. The two neural donor sources were then imported through the separate ordinary C1QA-reference route; existing historical cache copies were preserved.
The review uses the user's explicit ClinGen task instruction to retain supported generic protein binding as non-core when a more specific activity is not justified. This is a task-specific instruction, not fabricated maintainer approval. The current published project policy records this treatment. Qualified curator deference is disclosed per row and is not a claim that each screen's supplementary target table was read.
Human C1q consists of six A/B/C heterotrimers. C1QA contributes structural collagen-like and globular recognition elements; the C1r/C1s proteases carry out complement proteolysis. The human recombinant ghA abstract directly reports immunoglobulin binding: PMID:12847249(https://pubmed.ncbi.nlm.nih.gov/12847249/), “the MBP-ghA bound to heat-aggregated IgG and IgM”. The complete abstract, rather than full Methods, was read. PMID:19006321 also reports A-chain substitutions affecting IgG/IgM recognition. Physiological avidity and activation belong to the assembled C1q/C1 system, as supported by recombinant C1q assembly/activation (PMID:23650384), IgG-hexamer experiments (PMID:24626930), C1-IgG1 structures (PMID:29449492), and surface-binding/reconstitution work (PMID:34155115).
The core uses GO:0019865 immunoglobulin binding to synthesize the existing IgG- and IgM-binding siblings; it does not add a redundant broad NEW annotation. Official AmiGO GO:0019865 identifies IgG binding GO:0019864 and IgM binding GO:0001791 as direct children. The direct record initially timed out; the successful official indexed term record, with definition and parent/child relations, is retained separately. New core identifiers are checked by normal term validation.
The actual Methods/Results of PMID:18250442(https://pubmed.ncbi.nlm.nih.gov/18250442/) test human serum C1q and its globular region with apoptotic HeLa cells, phosphatidylserine-containing surfaces, cosedimentation, SPR and phosphoserine crystals. Its Discussion states “phosphoserine is bound to subunit C”. An A-chain contact is contributed by a neighboring crystal molecule and is packing-constrained. Thus the core describes C1QA as contributing to phosphatidylserine recognition within C1q, not possessing an independently established A-chain phosphatidylserine site. No NEW phagocytosis annotation is manufactured from a clearance or knockout phenotype.
The annotation consultant read actual human C1q/CD33 binding and ITIM-phosphorylation experiments in PMID:28325905. Only the existing CD33-specific protein-binding assertion is refined to GO:0048018 receptor ligand activity, already present as a separate IBA source row. The C1QB partner row from the same paper is not repurposed. Whole-C1q or heterotrimeric-fragment experiments do not establish isolated A-chain sufficiency. Activation of an inhibitory receptor is compatible with receptor ligand activity; it is not evidence of ligand activity with the wrong regulatory sign.
Direct ghA binding to mature gC1qR/C1QBP is supported in PMID:28018340 and as the actual positive control in PMID:34804054. The latter's viral title is not a reason to reject the C1QA interaction. The processed partner Q07021-PRO_0000018590 remains exact. C1QB association has independent heterotrimer-interface evidence from PMID:12960167. PMID:7240 establishes a covalent A–B disulfide linkage, which supports assembly but is not by itself a noncovalent protein-binding assay; that generic binding row is retained through curator deference and the independent interface evidence. The three UBQLN2/SGTB screen assertions retain their exact curated edges with framework-level assay reading and explicit uninspected individual target-table limits; no native ubiquitin-related activity is inferred. The 2017 erratum listed in the PMID:21988832 cache was not adjudicated.
The immutable PTHR15427 PAINT table was joined to the five existing IBA assertions. PTN008355511 carries extracellular location and receptor-ligand IBDs; PTN008559546 carries C1 complex membership, classical activation and synapse localization. These are inspected curator node placements, not reconstructed pairwise transfers. C1QA's occurrence among the experimental descendants for its own IBA is legitimate and is never called circular evidence. The full tree/MSA was not independently reconstructed and no residue-based divergence is asserted.
The official MGI C1qa homology GO graph exposed specific mouse/rat experimental provenance. Its displayed graph date is 2023-03-10, so it is not represented as an exhaustive current donor snapshot. It maps microglial/astrocyte activation to PMID:15269255; synaptic localization, complement-mediated pruning, neuron remodeling and eye-specific neural patterning to PMID:18083105; and several synapse-organization/pruning/postsynaptic assertions to PMID:27033548. The normal donor caches were subsequently read as detailed below. No missing fine subcompartment experiment is invented from the general synapse phenotype.
For PMID:15269255, the inspected mouse C1qa-knockout/APP-transgenic Methods and Results distinguish reduced microglial/GFAP responses and neuronal injury from unchanged amyloid burden. The A-chain itself is not claimed to form amyloid or act as an intracellular glial kinase. The existing contextual human orthology rows are preserved without claiming human interventions.
PMID:6249812 is bibliographic metadata only in the cache. Its two original IDA rows remain UNDECIDED, although modern independent sources establish the general biology. The normal PMID:28675934 cache supplies extraction/proteomics narrative but not the inspected C1QA-specific supplementary protein entry; its HDA matrix assertion remains UNDECIDED. Fine presynaptic/post-synaptic membrane and glutamatergic-synapse transfers are evaluated separately from general rodent synaptic C1q evidence. None of these limits implies an annotation is false.
Source-specific access notes are also recorded in every manually assessed PMID reference. Full-text availability and the sections actually read are separate facts. No figure-image or supplementary-table inspection is claimed when only a caption or narrative was accessible. Bibliographic TODO titles are reconciled only to the normal fetched cache title, preserving every original PMID and source annotation.
PMID:18083105(https://pubmed.ncbi.nlm.nih.gov/18083105/) is available as a complete normal cached abstract. It explicitly describes postnatal neuronal C1q at synapses, C1q/C3-deficient mice with failed retinogeniculate refinement and excess retinal innervation, and adult retinal relocalization in a glaucoma model. This supports the existing developmental mouse-donor assertions without establishing a human developmental experiment. The full 2007 Methods and the rat-specific culture experiment mapped by MGI were not inspected.
PMID:27033548(https://pubmed.ncbi.nlm.nih.gov/27033548/) supplies the full available main narrative. It reports structured-illumination microscopy of C1q with PSD95 puncta, mouse C1qa knockout and anti-C1q intervention, downstream C3 deposition, preserved LTP/synaptic puncta, and CR3-dependent microglial engulfment. These are distinct observations: C1q contributes recognition and complement initiation, while microglia execute engulfment. The complete cached main narrative was read; supplemental Methods, individual figure images and supplementary tables were not. General postsynaptic puncta colocalization does not by itself establish a membrane leaflet or each finer transferred synapse subtype, so three fine-location rows remain UNDECIDED.
ROOT's early review prompted two explicit integration refinements before sealing. The PMID:7240 row now separates covalent assembly from noncovalent binding evidence, as described above. The receptor-ligand IBA is ACCEPT, consistent with the independently supported contributed receptor-ligand core. CD33 in human monocytes is a demonstrated example, not a claim that CD33 is the dominant or universal receptor in every C1QA context. The binding consultant's original report and bounded core consultation remain unchanged and are credited as consultations, not whole-review approval.
No NEW assertion is proposed. The three cores synthesize immunoglobulin recognition, contributed phosphatidylserine recognition, and contributed receptor-ligand signaling; no complement protease activity is assigned to C1QA. No biological claim rests on counting PAINT donors. Status remains DRAFT because source-specific uncertainty and intentional generic-binding warnings are retained.
The ordinary verbose schema, term and best-practice validation passed on the isolated TMP candidate with 12 warnings: nine intentional retained generic-binding rows under the explicit task instruction; two source-specific ACCEPT/UNDECIDED differences caused by the metadata-only PMID:6249812; and one broad immunoglobulin core term synthesized from existing IgG/IgM siblings. The last warning does not justify a redundant NEW parent assertion. The 83 source objects remain exact, with 44 ACCEPT, 32 KEEP_AS_NON_CORE, one MODIFY and six UNDECIDED; there are no NEW assertions. All 38 original reference identifiers remain, three TODO titles are reconciled to actual normal cache titles, and the two donor papers plus the immutable UniProt file reference bring the total to 41. No alternative product, source field, cache or canonical curated file was changed in preparing this candidate.
The exact YAML and notes approved by the distinct internal ROOT whole-science peer were applied after checking source preimages. Normal canonical validation passed with the same 12 documented warnings; standard CREATE history and targeted rendering passed. All 83 source assertions and 38 original reference identifiers remain; the final 41 references include the two ordinary donor caches and UniProt file. No product slot or NEW assertion was introduced. DRAFT and six UNDECIDED rows remain. Immutable source/cache bytes were preserved. Remote publication is a separate ROOT-owned step.
C1QA supplies a recognition chain of C1q rather than a complement protease. Human recombinant ghA binds immunoglobulins (PMID:12847249); the globular heterotrimer has noncovalent interfaces (PMID:12960167), and assembled human C1q binds and activates the C1r/C1s protease assembly (PMID:23650384; PMID:34155115). The broad complement-activation assertion is correct, but its supported classical-pathway child GO:0006958 gives the more specific description. The official ontology places GO:0006958 directly under GO:0006956. The original TAS source and all machine annotation fields remain intact.
Other recognition roles have different experimental scopes. Phosphoserine occupies a chain-C site in the C1q head (PMID:18250442), so the A-chain contribution is expressed as part of the assembled recognition unit. Whole C1q and its globular region engage CD33 and activate its inhibitory signaling motifs in human monocytes (PMID:28325905); this does not establish isolated A-chain sufficiency. Recombinant ghA also binds mature gC1qR in the reported assays (PMID:28018340), and is the positive control in the Figure 5 binding assay of PMID:34804054. Ten short, exact cached-text anchors now document these claims and the principal neural observations in the review's reference findings; quotes are not repeated here.
The neural evidence remains explicitly rodent and contextual. The complete PMID:18083105 abstract describes postnatal synaptic localization and failed retinogeniculate refinement in complement-deficient mice; the full Methods are not available in that cache. PMID:27033548 directly reports C1q colocalization with PSD95-positive puncta in mouse hippocampus. These findings support the retained broad synaptic and postsynaptic annotations. An extrinsic membrane component need only associate with a membrane surface, so leaflet-level resolution is not a requirement. The exact donor experiments for presynaptic-membrane association, postsynaptic-membrane association and the glutamatergic-synapse transfer were not exposed in the inspected donor records. Those three source-specific decisions remain UNDECIDED, without rejecting synaptic C1q or treating the curator's inference as disproved.
Ontology access: classical complement activation, extrinsic presynaptic membrane component, and glutamatergic synapse. The direct postsynaptic-term page timed out; no successful reading of that page is claimed. Earlier dated journal entries are retained as historical records. The description now states biology without annotation-workflow commentary; the exact source fields, all three core functions, and supported non-core binding decisions are preserved.
Applied the independently reviewed evidence and specificity changes. Canonical validation passed with 12 explained advisories; targeted HTML rendering and the new EDIT history validation passed. All 83 source assertions, 41 references and three core functions remain. The ten exact quotations and their access limits are retained; DRAFT and six UNDECIDED decisions are unchanged. Earlier journal entries, history records and source files were preserved.