The authenticated primary catalog lists CFH unchecked. HGNC:4883 and reviewed UniProt P08603 identify the intended gene. All 67 normal source assertions, two alternative-product records and 44 original reference identifier/title pairs are preserved. This review adds no annotation. The current decisions are 19 ACCEPT, 13 KEEP_AS_NON_CORE, 2 MARK_AS_OVER_ANNOTATED, 28 MODIFY, 5 UNDECIDED.
The source packet and exact lookup scope are recorded in CFH-source-evidence.json. That file contains literal interaction identities, methods, taxa, roles, features and figure locators, plus current official term definitions and Reactome event records. It deliberately distinguishes a curated source join from a new experiment.
Human factor H recognizes C3b, supports factor-I-mediated C3b processing and accelerates alternative-pathway convertase decay. Factor I supplies the protease catalytic site; factor H performs noncatalytic cofactor work. The transient C3b–miniFH–FI structure uses engineered miniFH and FI-S525A, while separate biochemical assays establish cleavage. Full-length FH contains 20 CCP domains; FHL-1 shares the N-terminal seven and has a distinct short tail, without CCP19–20. The human FH/FHL-1 decay study is available as a genuine normal abstract, with its access limit retained. [PMID:28671664; PMID:21317894; PMID:8898949]
Host-polyanion recognition is a second molecular capability. Heparin experiments are informative biochemical models for heparan-sulfate recognition; they do not identify a particular proteoglycan core-protein interface. Full-length and truncated-product domain assignments remain distinct. [PMID:22471560; PMID:16612335]
Supported generic protein associations are retained as non-core under the standing explicit ClinGen project instruction. More informative binding terms are used where partner class and assay support suffice. No source annotation is deleted or silently rewritten. PAINT ancestral-node history is not inferred from donor count or target self-reference.
Normal fetch completed before manual review. Existing main publication, Reactome and family caches were selected byte-for-byte; fresh normal variants remain separate. Genuine context-complete Falcon research timed out, and the explicit fallback returned a quota error. No synthetic provider report was written. Two missing supplemental PMIDs were fetched through the normal CLI. Optional full-text availability flags are omitted where cache availability and actual external access differ; reference assessments state both.
Short supporting quotations occur only in the review YAML and are audited across the newly authored packet, including repeats. The authoring ceiling is 25 words per source; it is not described as a repository validator rule. No primary quotations are added to these notes. Raw downloaded documents and machine records remain separately identified archival evidence.
The mechanism/product peer independently verified performed cofactor work, exact C3 processed-product identities and FH/FHL-1 boundaries. The pentraxin peer independently checked ligand forms and exact source records. Whole-gene and publication review remain separate gates; this research packet does not itself mark the campaign gene complete.
The read ledger distinguishes selected full primary sections, actual figures/legends, abstract-only access and unresolved supplements. HTTP-success challenge pages were excluded. An early incorrect PMC resource request and an initial self-pair join overcount were explicitly corrected in research receipts before candidate authoring; neither altered any cached source or biological source object.
This entry supersedes the initial developmental uncertainty and refines the current core summary; the original notes above remain unchanged as a dated audit record.
The exact ligand in PMID:22471560 is heparan sulfate, including isolated glycan fragments PMID:22471560. Row 60 is therefore refined to GO:1904399 and the host-recognition core uses that MF. This does not assert that binding a heparan-sulfate proteoglycan requires a direct core-protein contact: binding its glycan moiety can also meet the broader definition. The narrower recommendation follows the assayed chemical entity. Existing heparin annotations remain correct with the analogue/sulfation caveat.
A separate C3d-binding core now records the directly resolved human FH19–20 recognition capability supported by PMID:21317894 and PMID:16612335. FHL-1 lacks those C-terminal domains. The first core retains C3b binding, noncatalytic factor-I cofactor work and alternative-pathway inhibition, but no longer repeats generic proteolysis or promotes the captured transient cleavage assembly to its defining complex. The original IDA proteolysis/complex assertions remain accepted: performed cofactor activity is real, and neither an intrinsic FH protease site nor a universally stable peptidase complex is claimed.
The newly recovered full PMID:30705315 changes the assessment of the two broad electronic developmental annotations. In neonatal Cfh-null mice, mitotic and pyknotic retinal profiles shift between birth and days 4–7; neuroblast dimensions and lamination also differ. The primary describes altered PMID:30705315. These are positive developmental findings, distinct from adult neuroprotection in PMID:42686909. Both existing assertions are retained as non-core: neurogenesis has direct cell-production timing context, while cell development is the narrower inference from precursor morphology and retinal maturation. Histological snapshots do not provide live lineage tracing or measured commitment, and no direct human developmental mechanism is established. Adult retinal thickness converges, and the RPE-to-precursor interaction remains unresolved. The weaker heterozygote figure comparisons are not turned into a strong haploinsufficiency claim. No developmental core or NEW process is introduced.
The neutral upstream-or-within IDA complement-regulation row and directional ARBA refinement reflect their different source scopes, not opposite biology. Independent direct FH evidence supports retention of complement regulation as core; an indirect source assay alone does not require downgrading an otherwise defining function. Pentraxin-binding refinements retain a family interpretation with monomeric CRP and long-pentraxin PTX3 forms stated explicitly; a suggested question now records the tension with the definition's pentamer wording. The description's neuronal-protection sentence is explicitly mouse-scoped.
The source-evidence JSON remains a literal machine-record/primary-access extract, with this update appended rather than rewriting raw records. It is not a custom bioinformatics computation. Its full authenticated historical data and prior record hashes are preserved. The published instruction distinguishes YAML from authored TSV; no new TSV or synthetic provider output is created. Source-field/product/ref identities and the complete published notes prefix remain protected. New-source quotations, including repeated YAML and this appendix, remain within 25 words per source; inherited unchanged notes are explicitly historical. Normal validation/status/rendering and exact-head guards are performed before ROOT publication.