CPAMD8

Summary: Human CPAMD8 supports normal camera-type eye development, especially anterior-segment formation.

Reason: PMID27839872 identifies recessive CPAMD8 variants in three human families with iris and lens dysgenesis and directly tests abnormal splicing in patient RNA or minigenes. Its developmental expression analysis is consistent with the affected structures. The current GO:0043010 definition specifies the camera-type eye and is a more precise child of the original eye-development term for this human evidence. Preserve the curated IMP participation claim without claiming that expression alone identifies the molecular step. Independent human and zebrafish studies support anterior-segment development, while the extracellular effector mechanism remains unresolved; no new developmental process is proposed.

Proposed replacements: camera-type eye development

Summary: The electronic endopeptidase-inhibitor assignment remains unresolved at target-specific activity level.

Reason: CPAMD8 itself has a predicted Kazal-like domain at residues 1706–1757 in the reviewed sequence, with IPR002350, PF07648 and SSF100895 domain-family assignments. This is independent architectural support for a possible protease-inhibitor activity and is distinct from the literal IEA sources IPR001599 and IPR002890, which are broad alpha-macroglobulin/MG2 mappings. The original PMID:15177561 abstract describes a serine-proteinase-inhibitor/follistatin-like Kazal domain and reports heparin binding; neither its domain label nor that ligand-binding result establishes the binding and reduction of endopeptidase activity required by GO:0004866. PMID:32274568 presents the inhibitor mechanism as putative. Full original biochemical Methods/Results remain unavailable, so a target-specific inhibition assay cannot be confirmed or excluded. Retain UNDECIDED: the architecture supports the hypothesis, but inhibition, a specific inhibited protease and any loss of inhibitory activity remain unresolved.

Summary: CPAMD8 has a soluble extracellular pool distinct from its membrane-associated pool.

Reason: CPAMD8 has a soluble extracellular pool: PMID:32274568 reports N- and C-terminal fragments in human aqueous and vitreous humor using three antibodies, separately from tissue immunofluorescence. Intact CPAMD8 was not identified in these fluids. This supports extracellular protein products without assigning a fragment-specific activity or cleavage mechanism. The current GO:0005576 definition accommodates this pool; cell-surface attachment alone would not. GO:0005615 extracellular space is obsolete and replaced by GO:0005576: fresh QuickGO and OLS responses on 2026-10-12 confirm this status, independently documented in the completed ontology issue [GO issue 31545](https://github.com/geneontology/go-ontology/issues/31545) and the [current OLS record](https://www.ebi.ac.uk/ols4/ontologies/go/classes?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FGO_0005615). Older repository label caches do not establish the current term status. The fluid location therefore does not justify replacing the source term with that obsolete ID. The literal InterPro and secretion-vocabulary mappings agree with the independent target evidence. Retain this basic secretory localization as core cellular context; the unresolved physiological molecular activity, represented by an empty formal core-function synthesis, does not negate the established location.

Summary: CPAMD8 can associate peripherally with the plasma membrane.

Reason: The original PMID:15177561 abstract reports ionic membrane association. Reviewed UniProt annotates peripheral cell-membrane association on the extracellular side with experimental support from that paper, and SL-0039 maps to GO:0005886. Retain the broad electronic location as a basic property of the membrane-associated CPAMD8 pool, compatible with separately detected soluble extracellular fragments. It is not merely a downstream developmental phenotype, so ACCEPT and NO_FAILURE_CORE describe the cellular location while the primary molecular effector remains unknown. No transmembrane helix, GPI anchor or additional surface-topology refinement is inferred; original assay and construct details remain unavailable.

Summary: The CPAMD8 Kazal domain has a reported heparin-binding activity.

Reason: The primary abstract explicitly reports Kazal-domain binding to heparin, and reviewed UniProt independently records the same paper as experimental interaction evidence. This narrow molecular-function proposal retains that positive result without claiming an endogenous heparan-sulfate ligand, a binding affinity, an isoform-specific assay or a causal heparin-binding step in eye development. Full Methods/Results and the experimental construct boundaries were not recovered. Current typed ancestry shows no ancestor/descendant redundancy with the existing GO:0004866 activity; the other original terms are BP or CC. It is not promoted to a formal core function because its physiological importance is unresolved.