FANCA (FANCA) — AIGR vs Affinage
Affinage record: run 2026-06-09 · 29 discoveries · self-eval win · gates passed.
Agreement (brief)
- Both place FANCA as a non-catalytic scaffold/organizer of the FA core complex whose
functional pool is nuclear, required for assembly, stability, nuclear accumulation and
chromatin loading of the complex that monoubiquitinates FANCD2–FANCI. - Both anchor the same partner network (FANCG, FANCF, FANCC, FAAP20) and the same core
process (interstrand cross-link repair / the FA–BRCA pathway), citing an overlapping set
of foundational PMIDs (9398857, 11063725, 22266823, 22343915). - Both agree the cytoplasmic pool is a minor/pre-assembly location, not the functional site.
Disagreements
| Topic | Affinage says | AIGR review says | Verdict (who is right + why) |
|---|---|---|---|
| Molecular activity grounding | mechanism_profile lists coarse MF parents: DNA binding, RNA binding, catalytic activity, acting on DNA (GO:0140097), molecular adaptor activity (GO:0060090), structural molecule activity (GO:0005198) |
Specific curated core MF: protein-macromolecule adaptor activity (GO:0030674) with the scaffold role in core_functions |
AIGR for the primary function. GO:0030674 is a precise child of the adaptor parent Affinage emits; structural molecule activity and the bare catalytic activity, acting on DNA are lossy/misleading down-casts. But Affinage's DNA/RNA-binding terms flagged a genuine gap (see below). |
| Intrinsic nucleic-acid binding | FANCA binds nucleic acids (RNA > ssDNA > dsDNA) via its C-terminal domain (PMID:22194614) | Not annotated anywhere in the review | Affinage surfaces a real gap. Verified in full text: purified FANCA, EMSA, patient truncation Q772X deficient. Incorporated as NEW single-stranded DNA binding (GO:0003697), non-core. |
| Single-strand annealing / strand exchange (RAD52-like) | FANCA directly catalyzes SA and SE comparable to RAD52 and acts in the SSA sub-pathway of DSB repair independently of the canonical FA pathway (PMID:30057198) | Not annotated; review covers only ICL repair / DNA repair | Real but bounded. Mol Cell 2018, purified protein + homodimer requirement + six deficient patient mutants + CRISPR/siRNA cell assays. Genuine and disease-relevant, yet single-lab (Zhang group) and not adopted by GO curators. Incorporated as NEW double-strand break repair via single-strand annealing (GO:0045002), explicitly non-core — not elevated to a core function. |
| RNA binding (GO:0003723) | Emitted as an MF term; narrative notes RNA affinity > DNA | Absent | Neither annotated. Kept out deliberately: even the source paper concedes "there is currently limited information to explain how the RNA-binding activity of FANCA could be linked to its functions." Documented in the reference notes, not annotated — a correct conservative call. |
| Centrosome / mitotic-spindle role | Narrative asserts FANCA localizes to centrosomes/PCM and maintains spindle-assembly-checkpoint function (PMID:23806870, PMID:26366677) | Not in the review | Left out (defensible). These are real papers but a peripheral, pleiotropic activity not in the curated GOA; consistent with AIGR's scope discipline. Not incorporated (conservative). |
| ATR-S1449 / NEK2-T351 phosphoregulation | Narrative highlights DNA-damage-induced ATR phosphorylation and NEK2 centrosomal phosphorylation | Not annotated | Regulation of FANCA, not a function of FANCA. Correctly excluded by AIGR — these describe upstream regulation, not a GO function/process/component for FANCA itself. |
protein binding (GO:0005515) IPI annotations |
Would collapse the whole hub into generic binding / adaptor | 18 GO:0005515 IPI calls MARK_AS_OVER_ANNOTATED; hub role captured once as GO:0030674 | AIGR — the curated treatment (informative adaptor MF + over-annotation flags) is strictly better than a generic-binding collapse. |
Papers incorporated into the review
| PMID | Supports | How used (supporting_text / reference / NEW annotation) |
|---|---|---|
| PMID:22194614 | Intrinsic ssDNA/nucleic-acid binding | Added to references (reference_review MEDIUM/VERIFIED) + NEW annotation single-stranded DNA binding (GO:0003697, IDA, non-core) with two verbatim supported_by quotes |
| PMID:30057198 | SA/SE (RAD52-like) role in SSA DSB repair | Added to references (reference_review MEDIUM/VERIFIED) + NEW annotation double-strand break repair via single-strand annealing (GO:0045002, IMP, non-core) with two verbatim supported_by quotes |
Not incorporated (deliberately): PMID:23806870, PMID:26366677 (centrosome/spindle), PMID:19109555
(ATR-S1449), PMID:24799500 (SHM/CSR), PMID:25015289 (CXCR5 neddylation), PMID:20864535 (NPMc) —
peripheral/pleiotropic or regulatory activities not in the curated GOA; excluding them is the
conservative, scope-appropriate call. Most of Affinage's remaining PMIDs are already cited by the
review.
Net assessment
The review is modestly better after reconciliation. AIGR was already stronger than Affinage on
the primary function (a precise adaptor MF and disciplined over-annotation handling versus Affinage's
coarse, sometimes wrong-branch mechanism_profile). Affinage's real value here was its dense
narrative, which surfaced one genuine gap the curated review had missed: FANCA's intrinsic
nucleic-acid binding and RAD52-like SA/SE activity in the SSA branch of DSB repair. These are
well-controlled (purified protein, disease-mutant correlation, cell assays) but single-lab and not
yet adopted by GO curators, so they were added as two NEW non-core annotations rather than
promoted to core functions — capturing the finding without overstating FANCA's canonical scaffold
identity.