BRCA2 (FANCD1) — AIGR vs Affinage
Affinage record: run 2026-06-09 · 27 discoveries · self-eval pairwise = win, faith 100% · gates passed.
The Affinage record for BRCA2 is a strong, PMID-dense (27 citations) mechanistic narrative of the
RAD51-mediator life cycle: C-terminal DBD (three OB folds + HTH) binding ssDNA, full-length BRCA2
selectively loading RAD51 onto ssDNA over dsDNA and displacing RPA, single-molecule acceleration of
RAD51 nucleation on RPA-coated ssDNA, BRC-repeat non-equivalence, TR2 filament brace, PALB2/BRCA1
recruitment, FA-pathway/FANCD2 chromatin loading, reversed-fork protection, R-loop control,
phosphoregulation (ATM/ATR-PP2A-B56, PLK1-T207), meiotic DMC1 mediation, and metabolite-triggered
(formaldehyde/methylglyoxal) proteolysis. The AIGR review is curated, GOA-grounded, validated, and
was already the product of a careful A→Z manual pass plus a 2026-07-22 QA re-review.
Agreement (brief)
The two sources agree on the core biology. Both center BRCA2 on homologous recombination as the
defining RAD51 mediator: ssDNA binding via the OB-fold DBD, loading/stabilizing RAD51 onto
RPA-coated resected ssDNA to build the presynaptic filament, D-loop/strand invasion, and reversed-
/stalled-fork protection from nucleolytic degradation (AIGR: GO:0003697 ACCEPT, GO:0000724 ACCEPT
across IBA/IEA/IDA/IMP, GO:0031297 ACCEPT, NEW GO:0000730 / GO:0042148, two core_functions; Affinage
narrative + PMID:12228710/20729832/36976771/29038466). Both treat nuclear/nucleoplasm/chromosome
localization, the BRCA1-PALB2-BRCA2 axis (GO:1990391 DNA repair complex), telomere and centrosome
pools, and meiotic HR (spermatogenesis, male meiosis I) as established, with the germline/meiotic and
centrosome roles demoted to non-core. Both also treat the intrinsic-HAT claim skeptically.
Disagreements
| Topic | Affinage says | AIGR review says | Verdict (who is right + why) |
|---|---|---|---|
| Mechanism-profile GO layer | molecular_activity = GO:0003677 DNA binding, GO:0140096 catalytic activity acting on a protein, GO:0060090 molecular adaptor activity, GO:0098772 molecular function regulator activity |
Core MF is the specific GO:0003697 ssDNA binding; BRCA2 is not a protein-acting enzyme, adaptor, or generic regulator |
AIGR right. Affinage's own header flags this layer as coarse/collapsed-to-parents and self-contradictory. catalytic activity, acting on a protein is wrong-branch (BRCA2 has no such catalytic activity); DNA binding is the less-specific parent of the annotated ssDNA binding. Not imported. |
protein binding (GO:0005515) |
Narrative leans on dozens of PPIs (RAD51, PALB2, DSS1, FANCD2, DMC1, p53, USP11, HSF2BP, MEILB2…) | ~40 IPI protein binding rows REMOVE'd as uninformative; interactors retained in summaries / specific-MF / complex terms |
AIGR right per repo-wide curation policy; Affinage does not assert the generic term, so no true conflict. |
| Meiotic DMC1 mediation | Distinct, well-cited function (PMID:26976601, 31242413, 32345962) | Captured only as non-core BP terms (spermatogenesis, male meiosis I, oocyte maturation, gonad development) | Both defensible; AIGR conservative-correct. DMC1-mediator activity is real but germline-restricted; demoting to non-core meiotic processes is appropriate. No specific mitotic-vs-meiotic MF gap forces a new term. Left unchanged. |
| R-loop suppression / RNase H2 recruitment | Real function (PMID:30560944, 28575672, 35715464) | Not annotated | Affinage right that it is real, but coarse/indirect for BRCA2. Left unannotated conservatively: BRCA2's R-loop effect is via recruiting RNase H2 and via fork stability, not a demonstrated intrinsic BRCA2 MF; adding a NEW term here would over-reach on this high-quality existing review. |
| Metabolite-triggered proteolysis, PLK1/PP2A phosphoregulation | Emphasized (PMID:28575672, 38608703, 32286328, 34593815) | Regulation of BRCA2, not BRCA2's own function; not annotated as BRCA2 GO | AIGR right to exclude. These describe upstream regulation/degradation of BRCA2, not molecular functions/processes executed by BRCA2. Correctly out of scope. |
| HAT activity | Not asserted (Affinage does not carry the disputed intrinsic-HAT claim) | GO:0010484/0010485 UNDECIDED; negated GO:0004402 (NOT HAT) ACCEPT | AIGR right / no conflict. Affinage's silence corroborates skepticism toward the contested 1998 intrinsic-HAT claim. |
Papers incorporated into the review
| PMID | Supports | How used |
|---|---|---|
| PMID:12228710 | ssDNA binding (GO:0003697) | Yang et al. 2002 Science crystal structure of the BRCA2 DBD-DSS1-ssDNA complex. Added verbatim supported_by ("demonstrate that this BRCA2 domain binds single-stranded DNA") to the GO:0003697 IBA annotation, which previously had only the falcon deep-research file + an indirect RAD51-recombination quote; reference + reference_review (HIGH/VERIFIED). |
| PMID:29038466 | Replication fork processing (GO:0031297) + core function 2 | Mijic et al. 2017 Nat Commun (reversed-fork protection from MRE11/PTIP/RAD52). Added verbatim supported_by ("protects stalled replication forks from nucleolytic degradation") to the GO:0031297 annotation and ("RAD51 nucleofilaments on regressed arms, to protect them from degradation") to the fork-protection core_functions entry — both previously falcon-only; reference_review HIGH/VERIFIED. |
| PMID:36976771 | RAD51 loading (GO:0000730 / GO:0000724) + core function 1 | Bell et al. 2023 PNAS single-molecule study. Added verbatim supported_by ("BRCA2 regulates recombination by initiating RAD51 filament formation") to the NEW GO:0000730 DNA recombinase assembly annotation and to the RAD51-loading core_functions entry; reference_review HIGH/VERIFIED. |
All three PMIDs fetched/cached via fetch-pmid; every supporting_text is a verbatim substring of
the cached abstract/full text (confirmed against the reference validator — the only errors reported
are the pre-existing benign offline-fetch failures for file:/GO_REF/Reactome references). No
new GO terms were minted: each addition strengthens an already-present, correctly-branched annotation
or core function. Zero NEW annotations were added — the review already carried NEW GO:0000730 and
GO:0042148, and the only functions Affinage surfaced beyond those (R-loop, phosphoregulation,
degradation) are either indirect/regulatory or out of scope for a BRCA2 MF/BP under a conservative bar.
Net assessment
The AIGR review and Affinage agree on BRCA2's core RAD51-mediator/HR and fork-protection biology.
Affinage's own GO/Reactome mechanism_profile is coarse and in places wrong-branch (a
catalytic activity, acting on a protein MF BRCA2 does not possess; DNA binding where the review
correctly uses ssDNA binding) and was correctly not imported. Affinage's genuine value here was its
dense, well-anchored narrative supplying three canonical primary papers — Yang 2002 (ssDNA-binding
structure), Mijic 2017 (reversed-fork protection), and Bell 2023 (single-molecule RAD51 loading on
RPA-coated ssDNA) — that the GOA-grounded review supported only through the falcon deep-research file
or an indirect quote. These were incorporated conservatively as added references + verbatim
supported_by on existing annotations and core functions, with no change to any existing curation
decision and no weakening from Affinage's coarse GO layer. File remains ✓ Valid.