CFAP418

Summary: Photoreceptor inner-segment localization is supported independently of disputed ciliary-base enrichment.

Reason: The original UniProt location mapping is retained. Mouse full-length C8ORF37 expressed in knockout photoreceptors and endogenous rat serial-section immunoblots support the photoreceptor cell body, including the inner segment, in PMID:29440555. Marmoset immunostaining in PMID:36233334 also detects inner-segment protein. These are ortholog and primate-retina observations, not a direct endogenous human retinal localization experiment; they support the inner-segment location without making exclusive ciliary-base localization the premise.

Summary: Broad cytoplasmic localization captures a supported functional compartment.

Reason: The source assertion is broader than ciliary-base localization but is not superseded by it. The original human hTERT-RPE1 study reports cytoplasmic C8orf37 before serum-starvation-associated ciliary-base staining. The later mouse tagged-protein and endogenous rat fractionation study supports broad photoreceptor cell-body distribution. Lipid-binding evidence and membrane-associated effects do not require an exclusively ciliary compartment. Keep the original IEA source unchanged.

Summary: Human CFAP418 co-purifies with CAPNS1 in the organelle interaction screen.

Reason: The exact human Q96NL8–P04632 pair is present in source-linked IntAct records EBI-12449215 and EBI-12449224 from PMID:27173435. These are N- and C-terminal tandem-tag affinity-purification observations, not proof of a purified binary interface or calpain regulation. The previous scaffold-protein-binding replacement does not follow from this partner or assay. Retain the supported generic association under the explicit ClinGen project instruction when no evidence-backed finer activity is established.

Summary: Human CFAP418 interacts with FAM161A in yeast and proximity assays.

Reason: PMID:36233334 tests full-length human proteins and fragments by yeast two-hybrid and tagged-protein proximity ligation in HEK293 cells, with additional marmoset-retina proximity evidence. The study does not report the co-immunoprecipitation claimed in the previous review. Its N-terminal 1–75 fragment shows some yeast self-activation, but stronger partner-dependent growth and full-length controls support the interaction. Neither this association nor the off-target historical bioinformatics report establishes CFAP418 scaffold activity or binding to a demonstrated signaling scaffold. Retain supported protein binding under the project instruction without transferring FAM161A microtubule activity to CFAP418.

Summary: The ortholog-supported photoreceptor inner-segment annotation is retained.

Reason: Current source tracing identifies mouse CFAP418 Q3UJP5 with an IDA inner-segment annotation from PMID:29440555. Its full Results show tagged mouse protein throughout the photoreceptor cell body except the outer segment and corroborating endogenous rat serial-section data. This supports the original human ISS assertion at inner-segment resolution. The source object is not rewritten to incorporate the newer metadata.

Summary: Outer-segment organization is the experimentally resolved morphogenetic phenotype.

Reason: The actual experimental donor is mouse CFAP418 Q3UJP5, with IMP evidence from PMID:29440555. Multiple knockout alleles lead to disorganized outer-segment membrane discs during development; connecting-cilium, basal-body and rootlet ultrastructure remain grossly intact. The previous review incorrectly attributed these knockout experiments to PMID:22177090. GO:0035845 captures the resolved outer-segment organization scope, as a semantic lateral refinement rather than an asserted GO descendant of GO:0008594. This existing source-supported process annotation does not by itself establish a structural scaffold, direct disc-building step or intraciliary motor role. The refined process is retained as a source-supported non-core developmental consequence, not assigned as the direct activity of the lipid-binding core. The separate redundant authored proposal is withdrawn.

Proposed replacements: photoreceptor cell outer segment organization

Summary: The original human cytoplasmic observation remains valid.

Reason: The actual PMID:22177090 full text reports cytoplasmic C8orf37 in hTERT-RPE1 cells, followed by ciliary-base enrichment after serum starvation. The two locations are compatible. Later mouse and rat evidence supports a broad cell-body distribution rather than replacing cytoplasm with a single ciliary site. The normal repository cache is abstract-only; the genuine fresh normal full variant and independently retrieved PMC body were inspected for this source-specific statement.

Summary: Ciliary-base localization is supported, with nonexclusive and antibody-dependent scope.

Reason: The original human hTERT-RPE1 study reports ciliary-base staining after serum starvation using a commercial antibody tested on bovine retinal extracts; mouse retinal staining is separate. PMID:29440555 later found nonspecific staining with the same commercial antibody in mouse knockout controls and no basal-body enrichment of tagged mouse protein, although that protein was present in the basal-body region. PMID:36233334 reports marmoset ciliary-base enrichment using another antibody with tagged-protein specificity controls. These findings limit exclusive or universally enriched localization, but do not erase the original human observation or establish absence at the base. Preserve the IDA assertion as non-core while using broader supported compartments for the lipid-binding core.

Summary: Conserved phosphatidic acid binding inferred from the directly tested mouse ortholog.

Reason: PMID:37971880 Figure 8A and supplemental Methods test full-length mouse CFAP418, not human protein: retinal NM_026005-derived constructs produce His- and GST-tagged proteins in bacteria, and both bind phosphatidic acid and cardiolipin on immobilized-lipid strips. The original construct chain is explicit in PMID:29440555. This human ISS proposal uses mouse UniProtKB:Q3UJP5 as the supporting ortholog; no direct human assay, tag-cleavage control, affinity constant or membrane-vesicle experiment is claimed. The two specific lipid terms describe different tested ligands of one phospholipid-binding core. They do not establish lipid transfer, sequestration, kinase activity or a performed intraciliary transport step.

Summary: Conserved cardiolipin binding inferred from the directly tested mouse ortholog.

Reason: PMID:37971880 Figure 8A and supplemental Methods test full-length mouse CFAP418, not human protein: retinal NM_026005-derived constructs produce His- and GST-tagged proteins in bacteria, and both bind phosphatidic acid and cardiolipin on immobilized-lipid strips. The original construct chain is explicit in PMID:29440555. This human ISS proposal uses mouse UniProtKB:Q3UJP5 as the supporting ortholog; no direct human assay, tag-cleavage control, affinity constant or membrane-vesicle experiment is claimed. The two specific lipid terms describe different tested ligands of one phospholipid-binding core. They do not establish lipid transfer, sequestration, kinase activity or a performed intraciliary transport step. The same paper reports reduced retinal PGS1 abundance at P5/P10 and distorted photoreceptor mitochondrial shape (Figure 4), providing physiological context for lipid homeostasis. Mitochondrial position and cristae appeared relatively normal, and retinal oxygen consumption was unchanged under the tested conditions. These observations do not establish CFAP418 mitochondrial localization or direct cardiolipin binding in vivo.