The 21 imported contexts are assessed using original experiments, target-specific database records and current ontology definitions. Source identifiers, evidence codes and the four source-described products are preserved. References are cited without new publication quotations under the campaign instructions.
The captured PANTHER v19 tree places human EYS PTN000621604 below PTN008602128, which carries the extracellular-region and extracellular-structural IBD assertions. The ungapped target MSA equals the displayed 3165-residue sequence; the fly donor FBgn0031414 is bridged to its actual UniProt record and its MSA also matches. No NOT lies on the EYS inheritance path. A human NOT on another branch belongs to LAMC2. The normal UniProt record lists PTHR12916, whereas the actual v19 lookup and target-containing tree identify PTHR24033. These versions remain distinct; the inference uses the demonstrated target, ancestor and sequences, not a guessed family name or donor count.
Cook et al., PMID:18297055 directly measured extracellular structural work in fly experiments. Secreted Spam/Eys coated Prominin-expressing Kc cells, protected their shape against osmotic and chemical challenges, and increased indentation resistance after actin disruption. The complete main Methods/Results and all captions were read; no pixels or supplements were inspected. This supports conserved structural activity, without claiming an isolated human EYS mechanics assay. Earlier donor papers PMID:17011488 and PMID:17036004 were inspected only at metadata/abstract level.
Yu et al., PMID:27737822 shows that zebrafish Eys loss disrupts cone ciliary-pocket architecture before progressive cone and later rod degeneration. The complete main Methods/Results/Discussion and captions were read, and Figure 2 pixels were inspected. Photoreceptor maintenance is thus a supported role; the replacement GO:0045494 is explicitly lateral to the broad nervous-system assertion, not an asserted descendant of it. Developmental growth concerns increasing size or mass during development and is a distinct endpoint. The broad anatomical-development prediction remains undecided because its rule is unavailable and the intended developmental role has not been resolved.
The human disease studies 18836446, 18976725, 20696082 and 22363543 establish EYS-related photoreceptor disease. They do not establish EYS participation in the series of events in which light is received and converted into a molecular signal. That specific detection term does not require each participant to bind a photon, but retinal degeneration and impaired electroretinograms alone do not identify participation in that series. The four imported detection rows are redirected to GO:0045494 photoreceptor cell maintenance. Their clinical evidence is retained alongside the independent structural experiments described above: the maintenance inference combines a human requirement with demonstrated conserved structural work, rather than deriving participation from disease alone. UniProt CC FUNCTION explicitly assigns maintenance of photoreceptor integrity from PMID:18836446. The replacement is lateral to light detection and does not claim that the human genetics papers directly measured extracellular mechanics.
The experiments retain their actual scope. Figure 2B of PMID:18836446 uses mature porcine retina, separately from human family data. PMID:18976725 discusses PROM1/CRB contacts and extracellular sequestration as models rather than measured interactions. PMID:20696082 predicts a truncation from p.Glu1836Ter without assaying that protein. In PMID:22363543, 18 patients carried variants classified as very likely pathogenic, but only nine had two detected mutant alleles. Historical transcript lengths and predicted disease products are not equated with all current products.
PMID:21826682(https://pubmed.ncbi.nlm.nih.gov/21826682/) reports a suggestive association with severe statin-induced myopathy and transcript observations. Its skeletal-muscle regeneration mechanism is proposed from domain similarities and possible Notch/agrin-related signaling; the operative Methods/Results do not measure regrowth of injured fibers after EYS perturbation. This specific process interpretation remains over-annotated. The historical 3196-residue retinal/spinal transcript and muscle 3-prime amplification remain separate observations.
Alfano et al., PMID:27846257 distinguishes endogenous human Y79 staining from transfected constructs. Endogenous EYS was enriched with CEP135 at centrosomes and acetylated tubulin in cilia. Macaca fascicularis retinal sections showed partial overlap with acetylated tubulin and RP1 along outer segments. The complete operative main text and captions were read; Figures 3 and 8 were visually inspected. These support axoneme-associated and outer-segment pools. Axoneme includes associated proteins; the annotation does not make EYS a microtubule motor or tubulin. Centrosomal enrichment in a retinoblastoma cell line is retained as a contextual, non-core location.
Yu et al. Figure 2 includes rhesus macaque retina and one poorly preserved human donor sample. Blocking/omission controls were performed in macaque. Human signal enclosed by a PNA-positive sheath does not distinguish extracellular matrix from the enclosed outer segment. The extracellular pocket-lumen placement is a model; the authors report unsuccessful specific immuno-EM. The UniProt record explicitly assigns secreted interphotoreceptor matrix with ECO:0000269/PMID:27737822, in addition to a hedged secretion Note. That primary curated assertion is positive evidence and was understated in the earlier rationale. Nevertheless, enclosure by PNA does not itself determine which side of the photoreceptor membrane contains EYS. The pocket-lumen model and separate Y79 intracellular observations leave precise human topology unresolved; unsuccessful immuno-EM is a reported limit, not a universal requirement. Broad conserved extracellular structural support therefore remains compatible with an UNDECIDED precise IPM assignment.
The HPA EYS record uses HPA027103. Its sperm subassay is rated uncertain and annotated flagellar centriole. The selected image was viewed: focal green lies near the head–flagellum junction, without ultrastructural or specificity proof. A separately viewed uncertain U2OS image has cytoplasmic/peripheral signal and is not a replication of the sperm experiment. The 67-residue antigen matches positions 795–861 of both long products and is absent from both short products. It neither distinguishes the long products nor reports all four. The focal signal is positive, but the source-rated uncertain assay does not settle target specificity or precise centriole placement. This row is now UNDECIDED; independent Y79 centrosome enrichment is a different cell/compartment context and does not independently verify the sperm result.
| Literal source product | Source name | Length |
|---|---|---|
| Q5T1H1-3, displayed | 4 | 3165 |
| Q5T1H1-1 | 1 | 3144 |
| Q5T1H1-2 | 3 | 594 |
| Q5T1H1-4 | 2 | 619 |
The 2016 study explicitly compares differing database nomenclatures. Its full-length 619- and 594-residue constructs, split 3144-residue fragments and endogenous staining are different experimental objects; a 3165-residue expression construct was unavailable. Functional or localization equivalence is not established. The actual UniProt record has a predicted Signal feature at 1–21 and Chain at 22–3165. These are not observed universal cleavage or secretion. The initial preflight projection omitted the literal Signal type even though the source contained it; the intake records that projection limitation without changing the source. Separate RefSeq/Ensembl records and conflicting intracellular/secretion models remain explicit.
The original urinary proteomics study, PMID:19056867, used eight pooled healthy donors and a low-density membrane fraction enriched in exosomes. The relocated NHLBI database explicitly lists EGFL11, NP_938024, one peptide and reference 2, matching this paper. Target-decoy filtering and additional single-peptide inspection were described; gene-level disambiguation permits splice-product sharing. The historical peptide sequence/spectrum was not recovered. Current NP_938024.1 matches the 594-residue product, but does not identify the current product in that historical sample. The imported HDA location remains non-core, without universal secretion, vesicle purity or a core renal-function claim. The raw part_of UBERON:0001088 extension remains in the source archive; it was not manually inserted into the normal seed.
MatrisomeDB 2.0, PMID:36399478 combines sequence classification with curated proteomics, including enriched and soluble fractions. The authors' official human matrisome page links the revision-2014 masterlist; row 44 identifies EYS/HGNC:21555/Q5T1H1 as a core-matrisome ECM glycoprotein. This does not independently identify non-collagenous interstitial matrix. The precise tissue/peptide record remains unavailable: access failed TLS verification and the author notice reports retirement pending a new release, with a displayed date that is not silently corrected. A supplement request returned browser-check HTML, not an inspected archive. The interstitial assertion remains undecided rather than disproved. The captured UniProt record instead carries GO:0140144, non-collagenous component of basement membrane, while the imported GOA row carries GO:0140149, non-collagenous component of interstitial matrix. Official definitions confirm distinct compartments. The source versions and unavailable target dataset do not establish which location is correct or prove a specific upstream error; both literal records remain unchanged.
InterPro IPR001881 and IPR018097 currently map to calcium ion binding; the target has 20 matching domain intervals and five conserved-site intervals. This predicts a structural binding facet, retained as non-core, without a measured human affinity or a separate catalytic mechanism. The complete long-product architecture is not assigned to the short products. The five source location mappings were checked: SL-0458 outer segment, SL-0048 centrosome, SL-0066 cilium, SL-0304 axoneme and SL-0146 interphotoreceptor matrix. Mapping identity is separate from biological adjudication.
| ARBA source | Actual source result | Treatment |
|---|---|---|
| ARBA00029228 | Captured target FunFam condition matches | Cytoskeleton refined to supported axoneme association. |
| ARBA00027396 | Current rule retrieved; target match not reproduced | ECM retained on independent PAINT/original evidence; rule-version limit preserved. |
| ARBA00044390 | FunFam/taxon conditions match | Developmental growth marked over-annotated relative to maintenance evidence. |
| ARBA00029146 | FunFam/Metazoa conditions match | Lateral refinement to photoreceptor maintenance. |
| ARBA00098139 | Exact current resource returns HTTP 404 | Anatomical development undecided; no invented rule conditions or biological-loss inference. |
The seven publication-indexed original assessments used actual main text at recorded selective/full scopes, not the cache flag as a proxy for access. PMID:18976725 and PMID:19056867 have abstract-only normal caches but separately recovered full originals. For those seven papers, captions were read where recorded, but no figure pixels or supplements were inspected; complete variant tables in PMID:22363543 and the dataset table in PMID:36399478 were not fully reviewed. For the three independent experimental references, only the pixels explicitly listed above were viewed. No additional source reading is implied by combining the assessments.
All normal caches remain untouched. PMID:27737822 and PMID:27846257 were already genuinely fetched during intake; PMID:18297055 was fetched additively while protecting all preexisting runtime bytes. Ten cited PMID caches, two unchanged machine-associated PANTHER auxiliary files and both gene machine files form the proposed delivery closure. Six uncited acquisition caches remain private. The failed research-provider attempts produced no provider report or authored provider-branded substitute.
The single core combines the accepted extracellular structural inference with the photoreceptor-maintenance role. Direct evidence that Eys does structural work comes from the fly coating, deformation and indentation experiments in PMID:18297055(https://pmc.ncbi.nlm.nih.gov/articles/PMC2387185/). The zebrafish cone ciliary-pocket disruption preceding degeneration in PMID:27737822(https://pmc.ncbi.nlm.nih.gov/articles/PMC5155541/) supports a structural contribution to vertebrate photoreceptor maintenance. This is a conserved synthesis, not a human purified-protein mechanics measurement or an attribution based only on loss of survival.
The core's broad extracellular-matrix location follows that structural unit. It does not convert unresolved human interphotoreceptor-matrix placement into a resolved compartment. The native primate outer-segment and human Y79 axoneme/centrosomal observations remain positively supported and are retained as contextual, non-core locations. Their precise membrane topology and relationship to structural work are open questions, so they are not added as locations of the single extracellular structural unit. The source nomenclature, fragment experiments and HPA antigen do not establish that all four products share activity, localization or processing.
Only one BP is used, GO:0045494 photoreceptor cell maintenance, whose definition concerns preventing degeneration. No sensory light-detection step, upstream parent BP or independent second activity is added. Current GO:0005201 explicitly describes structural integrity of extracellular matrix, and the captured GO:0031012 record links that MF through occurs_in. No protein-containing complex is placed in the location slot. The questions propose future experiments; their hypotheses are not treated as established annotations.
The successor rereads the cited human genetics/clinical Results and Methods, the fly structural assays, the vertebrate ciliary-pocket and localization passages, the full UniProt location assertion and the selected HPA sperm image. No new original acquisition, peptide identification or retinal figure-pixel reading is claimed for this phase. The four maintenance replacements reuse existing references and the accepted structural participation model. The regeneration term remains over-annotated without a separate evidence-code defect claim; the two matrisome locations remain a recorded source-version discrepancy rather than an invented corrected annotation. The selected public caches for PMID:18976725 and PMID:19056867 are abstract-only, so their full_text_unavailable flags are true; this describes cache availability and does not negate independently recovered full-original verification. All original source bytes and the four literal products remain intact.