Author: Codex, /root/bloc_followup. The 39 partner-specific binding decisions were consulted with /root/bicra_science; 21 inherited/orthology decisions and two correction-access checks were consulted with /root/bloc_followup/brca1_late_binding. These are disclosed bounded consultations, not whole-gene approval. ROOT performs the separate whole-science review before any canonical application.
The existing normal UniProt and GOA files identify human C1QBP, Q07021, HGNC:1243, GeneID:708; aliases include GC1QBP, HABP1, SF2P32, p32 and gC1qR. The mature processed chain starts at residue 74 and extends to residue 282. This processing is not an alternative isoform. Neither products nor alternative_products is present in the machine review, and no product objects have been invented. Viral processed partners and the CDK13 partner isoform remain exactly as supplied in GOA.
The authentic Falcon/Edison research file already existed, with a January 18, 2026 run recorded in its metadata. It is preserved unchanged. No replacement provider output was synthesized and no redundant new provider call was made. The current decisions rely on normal cached primary sources and explicit access checks rather than unverified research prose. The unresolvable authored reference deep-research, whose title incorrectly described a 2024 Falcon summary, is removed as a duplicate; the actual file:human/C1QBP/C1QBP-deep-research-falcon.md reference remains. This removes no GOA source identifier.
The original review contained 118 annotation objects, whereas the preserved GOA file contains 125 records. A normal, TMP-only seed-goa --no-fetch-titles run restored seven omitted partner/product distinctions and populated source qualifiers and supporting entities. Every raw record now has an exact source object. None is merely a date-only duplicate, and none is a new biological assertion manufactured by the reviewer.
The restored distinctions are PRKCZ versus PRKD1 (historical PKC-mu), CDK13 versus CDK13-2, mouse versus human ARF, APOA1 versus LBR, VTN versus YBX1, YBX1 versus LBR in a second source, and factor XII versus kininogen. The source partner Q6RFH8 currently has the UniProt DUX4L9 name while the inspected papers use DUX4c; its accession is retained without substitution for a canonical DUX4 protein. POLGARF is kept distinct from POLG and from the released POLGARF fragment. No target NOT or target-isoform fields were introduced.
Four distinct, existing molecular activities organize the review. They are not four names for one interaction.
C1QBP is predominantly mitochondrial but not universally mitochondria-only. Later cell/tissue localization, surface antibody detection and purified endothelial p33 reconcile the matrix and surface pools. Rat tissue observations are not renamed as human assays. Nuclear/cytoplasmic transcription, splicing, preribosome remodeling, DNA-damage and apoptosis functions remain contextual. The Hrk perturbation evidence supports an apoptotic role without establishing an autonomous p32 pore. The MRE11 paper explicitly supports deoxyribonuclease inhibition before ATM-dependent complex release; its repair-promoting title does not negate this inhibitory phase. GO:0060703 is a verified refinement of generic enzyme inhibition, not nuclease activity assigned to C1QBP.
The published ClinGen curation instruction records the user's explicit task-specific direction to retain supported, biologically correct generic protein binding as non-core when no justified finer activity is available. This deliberately differs from the skill's informational-exclusion recommendation; it is not a new repository-wide policy or external maintainer sign-off. Ten generic rows are refined to experimentally supported partner classes, including the explicitly tested APOA1 and high-molecular-weight kininogen ligands. These last two refinements supersede the original bounded binding consultation and are recorded in its separate amendment. Twenty-seven remain non-core with source-specific evidence or exact curated-edge deference. Two O00221 screen rows remain unresolved because the target assays were not inspected and independent target corroboration was not established. An abstract emphasizing a different protein does not establish wrong-gene curation. Screen-framework access is not mislabeled as inspection of every target supplement, and co-purification/proximity is not uniformly called purified binary binding.
For IBA, a target appearing in the donor list is legitimate experimental grounding. Short donor lists are not considered weak. The source PTN objects and PTHR10826 family identity are retained, but the actual PAINT IBD tree/MSA was unavailable locally, so independent node-placement verification is not claimed. Conserved functions are supported by independent human evidence and donor sources where inspected. The rat donor identity is verified; the exact donor experiments for presynaptic active zone, glutamatergic synapse and GABA-ergic synapse remain unresolved. General neuronal or mitochondrial localization cannot substitute for those fine locations.
Four additional normal donor caches were imported exclusively by ROOT and are now used: PMID:10409668(https://pubmed.ncbi.nlm.nih.gov/10409668/), PMID:11698413(https://pubmed.ncbi.nlm.nih.gov/11698413/), PMID:22904065(https://pubmed.ncbi.nlm.nih.gov/22904065/) and PMID:9191880(https://pubmed.ncbi.nlm.nih.gov/9191880/). The first, second and fourth are abstract-only; the third contains full text. The PKC paper's machine title says “ray,” while its abstract explicitly identifies rat hepatocytes; the source title is preserved. The release paper distinguishes Raji-cell release from peripheral lymphocytes under the particular short-incubation conditions, rather than proving universal secretion.
The official correction to the IL-12/PI3K paper states that Figure 3c repeated the TNF-alpha graph in the IL-6 position and supplies a replacement. The complete short notice was read through the indexed publisher result; the corrected image/data were not independently inspected. The notice does not state withdrawal of the IL-12/PI3K finding. The source-specific judgments retain that boundary. The liver-interactome erratum and exact O00221 target supplement remain unadjudicated.
The TMP candidate contains 125 assessed source objects: 31 ACCEPT, 74 KEEP_AS_NON_CORE, 15 MODIFY and five UNDECIDED; no NEW and no source deletion. It has 72 reference records and four core units. Ten short supporting quotations are exact normal-cache substrings, with at most ten quoted words per source across the candidate. This notes entry adds no source quotations. Other evidence is reference-only, not fabricated or paraphrased inside quotation fields.
Normal schema/reference/ontology/best-practice validation passed with 28 warnings: 27 retained generic-binding warnings under the explicit task instruction and one optional notice that the preserved research file is not used as annotation evidence. The source-specific primary support is intentional. The review is DRAFT, matching assessed content with warnings; it does not claim all unresolved evidence is settled or campaign completion. Canonical sources, original review/HTML, existing research and prior history remain unchanged at proposal time.
Applied the exact YAML and notes approved by the distinct internal whole-science review. Canonical validation passed with 28 explained advisories, the targeted status dry-run retained DRAFT, and focused HTML rendering and standard EDIT-history validation passed. All 125 source objects, 72 references and four core units match the approved candidate. The 66 protected raw, publication, Reactome and provider files remain exact. No source file was overwritten, no earlier history changed and no remote mutation was performed in this application.
The exact-head review identified three supported refinements. The two generic membrane assertions now propose cell surface (GO:0009986). Complete cached abstracts for PMID:8662673(https://pubmed.ncbi.nlm.nih.gov/8662673/) and PMID:11290596(https://pubmed.ncbi.nlm.nih.gov/11290596/) establish endothelial surface binding sites or surface detection. The official cell-surface definition permits externally attached proteins; no transmembrane segment or constitutive CK1/uPAR/gC1qR complex is inferred. These are refinements of existing assertions, not additional annotations.
The factor XII interaction now proposes protease binding (GO:0002020), consistent with the MMP14 partner-class refinement. The complete PMID:8710908(https://pubmed.ncbi.nlm.nih.gov/8710908/) abstract explicitly reports factor XII binding to isolated and recombinant gC1qR. The official definition asserts the partner class, not catalytic regulation. The earlier activation-state caveat does not justify retaining a generic protein-binding term. Activation state remains unestablished here; neither protease activity nor activation/inhibition of factor XII is assigned to C1QBP. The kininogen partner remains a separate, unchanged source object.
The optional receptor-subtype suggestion was also resolvable without new experiments or references. AmiGO's adrenergic-binding hierarchy and the official FlyBase term record verify alpha-1B adrenergic receptor binding (GO:0031692). The complete PMID:10409668(https://pubmed.ncbi.nlm.nih.gov/10409668/) abstract supplies the rat-library, receptor-tail mapping and transfected COS-7 association evidence. Both existing rat-derived assertions propose that subtype while retaining their IEA/ISS codes and donor identifiers. This does not claim a newly tested human interaction or add a receptor-signaling core. Direct AmiGO child-page access failed; its parent hierarchy and FlyBase definition were available.
The three fine-synaptic assertions remain UNDECIDED. The exact donor experiments and the basis for the activity-location qualifier have not been recovered. A hypothetical proteomics origin is insufficient to relabel them as over-annotations, just as mitochondrial matrix localization alone cannot rule out another pool. The existing expert question preserves that uncertainty. The four core descriptions already lead with the molecular mechanism and retain necessary species, preparation and catalytic-attribution limits; they are unchanged.
This focused proposal changes only five annotation reviews: 29 ACCEPT, 71 KEEP_AS_NON_CORE, 20 MODIFY and five UNDECIDED. All 125 source objects, 72 reference records, product fields, ten existing quotations and four cores are preserved. No source cache, provider output, earlier note or history record is rewritten, and no NEW assertion or additional quotation is introduced. The four source papers were read as complete cached abstracts for this follow-up; no new full-paper access is claimed. Normal validation is recorded separately before application.
Normal full schema, reference, ontology and best-practice validation passed for this proposal with 27 advisories: 26 retained generic-binding rows under the standing project instruction and one optional unused-research-file notice. DRAFT remains appropriate; the five unresolved source assertions are unchanged.
Applied the five reviewed specificity refinements after distinct internal ROOT approval. Normal canonical validation passed with 27 explained advisories, the status dry-run retained DRAFT, and focused rendering and standard EDIT-history validation passed. All 125 source assertions, 72 references, product fields, four cores and ten quotations match the approved candidate. Source caches and the previous history record remain unchanged. External PR review and CI remain separate publication gates.