DPH1 is the human OVCA1/DPH2L gene (HGNC:3003, Q9BZG8). OVCA2 is a separate gene. The review preserves all 21 seeded source objects and every imported relationship, including the single contributes_to activity assertion. There are 12 ACCEPT, 7 MODIFY and 2 KEEP_AS_NON_CORE decisions, with no additional annotation. Broad complex and location refinements retain the original assay resolution in their reasons. All existing source and family/module caches remain unchanged.
The human DPH1 study complements deficient MCF7 cells with CMV-driven wild-type or variant DPH1 and measures toxin-dependent labeling of the EEF2 diphthamide endpoint in extracts. Main Methods, functional Results, Discussion and Figure 1 are available separately from the abstract-only normal cache. These experiments establish human function without constituting a purified human ACP-transfer assay. Supplement downloads failed. The later correction changes a clinical Table 1 epilepsy frequency, not the reported functional experiment.
The subunit-resolved biochemical study tests yeast Dph1–Dph2 heterodimers, cysteine mutants, SAM-to-EF2 transfer and reductant dependence. Its Methods/Results and Figure 6 distinguish Dph1 cluster catalysis from the requirement for the Dph2 cluster under the Dph3/Cbr1/NADH reducing system. One integrated core therefore uses contributes_to_molecular_function and the existing synthase-complex term. All imported enables and contributes_to relationships remain literal; the reasons explain the complex requirement. Fe–S binding is a cofactor feature of this unit, not a second catalytic core.
The Reactome reaction places a DPH1-containing complex in cytosol but labels DPH2 as the active unit using archaeal homology. An archaeal enzyme named Dph2 does not establish which human subunit performs catalysis. This model attribution is kept separate from its positively supported target membership and location. The exact chain is reaction 5358494 → catalyst 5358378 → complex 5358365 → DPH1 component 5358384, residues 1–438.
The original 1999 paper, Subcellular Fractionation, Subcellular Localization and Figure 4C caption, supports nuclear and cytoplasmic OVCA1. Human ovarian epithelial fractionation detects the approximately 50-kDa species in both pools. Tagged COS-1 and human HOSE experiments show diffuse staining and perinuclear puncta. The indexed original text was inspected; figure pixels were not. Neither the p70/p85 immunoreactive species nor the historical 443-aa description is automatically equated with every current natural product.
The Human Protein Atlas independently reports nucleoplasm in A-549 and U2OS with HPA069750, at approved reliability and supported subassay validation. Nucleus is refined to nucleoplasm at gene level, preserving the older experiment's broader resolution. Cytoplasm is refined to the positively supported cytosolic pool. Nucleoplasmic and perinuclear locations remain non-core because their local molecular roles are unresolved, not because the observations are rejected. Other HPA cell-junction staining is documented without manufacturing another annotation.
The Rolland binary interaction study supplies reproducibility and identity criteria in Results and Experimental Procedures. The literal DPH1–TFCP2 pair has three positive, nonexpanded IntAct records: EBI-10456463, EBI-10545280 and EBI-10303209. Their reference-sequence checksums match Q9BZG8-4 (438 aa) and Q12800 (502 aa). The official TFCP2 record establishes its DNA-binding transcription-factor class. GO:0140297 describes binding to that class; it does not transfer transcriptional activity or establish a native transcriptional partnership for DPH1. The older RBM8 interaction paper is abstract-only and does not establish DPH1 RNA binding.
The PAINT IBD at PTN000079945 carries GO:0017183; the current PTHR10762 tree contains the human target below that node. A broad ancestral process shared by DPH1 and DPH2 is consistent with their joint chemical work. No donor-count or self-circularity objection is used. Current exact yeast P40487 and mouse Q5NCQ5 donor records, the InterPro IPR016435 target match and its literal GO mappings are recorded in the evidence guide. An unresolved independent UniPathway endpoint is not treated as invalidating convergent, checked support.
Official products are Q9BZG8-4/-1/-2/-5, of 438/443/363/215 residues. The 443-aa product adds five N-terminal residues; historical variant coordinates match that numbering, without authenticating every experimental ORF. The shorter products are natural sequence alternatives, not the clinical mutant constructs. Retention of annotated cluster-binding residues alone does not establish activity.
Normal fetching and publication caching succeeded in an isolated workspace. Falcon authentication failed, and the configured perplexity-lite fallback returned a quota error; no provider report was invented. Existing caches were protected, including their literal full-text flags. The separately fetched PMID:31463593 is identical to main and the pending yeast-review copy; PMID:16625196 is identical to the separately pinned pending cache. Ordinary final validation, rendering and history are reserved for canonical publication. The evidence guide provides public source locators and the actual access limits.
The DPH1–TFCP2 refinement remains a binding-capability assertion. GO:0140297 specifies the partner's DNA-binding transcription-factor class; it does not require DPH1 to regulate transcription. Three positive records from the same 2014 interactome study include validated two-hybrid EBI-10545280, alongside array and prey-pooling records. Both participants are human and the assay host is yeast. The original pipeline used identity-confirmed, reproducible pairwise retests; pair-specific orthogonal validation and an endogenous physiological role remain unestablished. Reference-sequence checksums support Q9BZG8-4 and Q12800, without independently authenticating an archived experimental insert or extending the finding to all DPH1 products. The existing endogenous-interaction question is retained, and no transcriptional process or binding core is added. The scoped ClinGen curation instruction supports an informative evidence-backed refinement and preserves supported generic binding when no such refinement exists.
The tumor-suppressor annotation associated with the historical name OVCA1 merits an explicit research question. PMID:10519411(https://pubmed.ncbi.nlm.nih.gov/10519411/) reports growth suppression, G1 accumulation and cyclin-D1 changes in human A2780 cells expressing OVCA1. These observations do not establish whether that phenotype is separable from diphthamide synthesis; that relationship is now posed as a question, without a new process annotation or separate core. The normal publication cache remains abstract-only. Selected indexed original passages supplement the accessible abstract but do not constitute a complete original or figure-pixel read.
This follow-up changes one annotation rationale and appends one question. All imported source fields, actions, replacements, products, core chemistry and prior notes remain intact. No new publication quotation is introduced; the two existing excerpts are retained unchanged.