DRC4 is the current human symbol for HGNC:4166/O95995; GAS8 and GAS11 are previous names. The qualified main tree contained no review or history under any of those names, so this is a new DRC4 review rather than a rename. Ordinary isolated fetching produced 170 source assertions from 171 raw GOA contexts. The duplicate context is the same axonemal IDA assertion from PMID:26387594 supplied by MGI and UniProt. Every seeded source dictionary, qualifier and supporting entity is preserved. The decisions are 32 ACCEPT, 40 MODIFY, 97 KEEP_AS_NON_CORE and 1 UNDECIDED; no NEW annotation is proposed.
The independent core consultation and the annotation review converge on one N-DRC scaffold mechanism. Human loss of DRC4 removes DRC3 from axonemes, disrupts doublet alignment and reduces bending amplitude while some dynein-arm markers remain present. Detailed scaffold geometry is resolved in algae; the human respiratory reconstruction identifies the homologous subunits and places predicted models into density. These are complementary evidence levels, not interchangeable human atomic-interface experiments. PMID:26387594(https://pubmed.ncbi.nlm.nih.gov/26387594/), PMID:37258679(https://pubmed.ncbi.nlm.nih.gov/37258679/).
The integrated core records scaffold activity, axoneme assembly, cilium movement, its regulation and DRC3 localization/retention. These processes are not redundant in the current typed GO closure. Axoneme assembly is below microtubule cytoskeleton organization; flagellated sperm motility is below cilium movement. Thus the broader organization and specific sperm context are explained within this unit without adding another activity or duplicate process entry. Axonemal anatomy remains a location, separately from membership in the N-DRC/axonemal nexin-link complex. None of this implies motor ATPase activity or loss of every dynein arm.
Conserved microtubule binding supports the scaffold mechanism. The 2007 source reports single-particle binding and GMAD/IMAD effects, but its full construct Methods were not recovered. The 2000 abstract explicitly identifies a human GAS11 fragment associated with microtubules; a title foregrounding trypanosomes does not negate that experiment. The 2006 original distinguishes human tagged protein and human cell/airway detection from purified mouse protein tested with rodent Golgi membranes. These sources support binding and accessory localization without assigning every assay to a current human splice product. PMID:17366626(https://pubmed.ncbi.nlm.nih.gov/17366626/), PMID:10969087(https://pubmed.ncbi.nlm.nih.gov/10969087/), PMID:16643277(https://doi.org/10.1111/j.1600-0854.2006.00411.x).
The Smoothened evidence remains positive and context-dependent. Human C-terminal GAS8 fragments interacted with human SMO in yeast two-hybrid assays. The subsequent co-immunoprecipitation and cellular experiments used mouse Gas8, including an engineered GMAD–CT construct whose GRK2-dependent signaling effect was significant; the full-length construct's effects were variable and nonsignificant. Mouse knockdown and zebrafish rescue experiments support the stated signaling context. A later mouse gene-trap study found motile-ciliary defects but normal tested SAG-induced Smo accumulation and developmental patterning. Its residual N-terminal Gas8–beta-geo fusion prevents an assumption of a clean protein-null comparison. Neither study erases the other, and no separate universal signaling core or kinase/GEF activity is asserted. PMID:21659505(https://pmc.ncbi.nlm.nih.gov/articles/PMC3149358/), PMID:27472056(https://pubmed.ncbi.nlm.nih.gov/27472056/).
Current HPA ICC/IF records support nonexclusive plasma-membrane, Golgi, cytosolic, primary-ciliary and sperm-segment pools. The cytoplasm refinements use independent human cytosolic records and do not pretend the original mouse assays resolved cytosol. The sperm localization records and a human infertility original support the flagellar context. Loss of a C-terminal antibody signal in that patient cannot distinguish absent protein from a truncated product, and the reported splice effect was predicted rather than measured in patient RNA. HPA GAS8/DRC4, PMID:37950557(https://pubmed.ncbi.nlm.nih.gov/37950557/).
All 109 HuRI partner assertions have literal IntAct records and matching parent-sequence checksums. Nineteen partners also have separately recorded sufficient-binding-region features; a parent checksum does not erase those fragment boundaries. The KRT27 BioPlex association has independent HuRI corroboration. The interaction extract is factual, while the review supplies the biological judgment. Positive pairs without an established narrower class remain non-core under the authorized project binding policy, not false positives. Binary interaction plus a verified class supports the bounded refinements; the partner's catalytic, transcriptional or cytoskeletal function is never transferred to DRC4.
The four final class corrections make this distinction explicit. TCEA2 assists elongation and stalled-polymerase recovery, EAF1 stimulates ELL/ELL2, and HEXIM2 sequesters P-TEFb. Those functions do not establish the DNA-bound transcription-factor/coregulator relationship required by the current coregulator terms, so these pairs remain positive non-core interactions. CDC37 instead has an established client-binding co-chaperone role and is directly observed contacting an unfolded kinase intermediate. Its verified DRC4 binary pair supports protein-folding chaperone binding without requiring DRC4 to be a folding client or a chaperone itself. The exact current definitions are preserved in the ontology response. CDC37 official record, PMID:27339980(https://pubmed.ncbi.nlm.nih.gov/27339980/).
Seven IBA assertions were traced to literal IBD records and the target's ancestral path in PTHR31543. Their descendant evidence includes human, mouse, worm and fish contexts as applicable. A target appearing among the descendants is expected and is not circularity; donor count is not used to grade the phylogenetic judgment. The positive mouse sources for microtubule binding, Rab3B association and sperm/ciliary localization remain species-scoped. The single unresolved annotation is negative regulation of proliferation from PMID:9790751: its authentic human GAS11 abstract is relevant, but target-specific full Methods were unavailable, so no false-experiment or wrong-gene claim is made.
The two natural products are O95995-1 (478 residues) and O95995-2 (453 residues, deletion of residues 1–25). Both sequences were independently reconstructed. Antibody detection of two forms does not establish functional equivalence, and historical fragments or engineered mutants are not relabeled as natural isoforms.
Provider acquisition was genuine: Falcon failed authentication and the configured perplexity-lite fallback returned insufficient quota. No provider report was fabricated. Existing publication caches retain their original bytes and flags; separately recovered originals and new ordinary cache outputs are distinguished in the source guide. Only two short new publication excerpts are reproduced, once each in the YAML: seven words from PMID:26387594 and eight from PMID:32296183. All other support is reference-only or paraphrased with locators.
The portable evidence package regenerates pair/construct, HPA and PAINT projections from genuine source files after checksum verification. It neither generates annotation verdicts nor treats source records as proof of native physiology. Publication validation, history and current-main qualification remain separate from these scientific notes; no campaign completion is claimed before merge.
The review preserves all 170 imported source assertions and both natural products. Twenty review entries are revised and two supported assertions are proposed: protein complex scaffold activity (GO:0140378) and regulation of cilium movement (GO:0003352). This addendum supersedes the affected earlier dispositions without rewriting their history. No new source quotations are reproduced.
The mouse airway/ependymal experiment in PMID:27472056(https://pubmed.ncbi.nlm.nih.gov/27472056/) supports cilium movement, not displacement of epithelial cells and not a substituted sperm endpoint. Results and Figure 4 measure waveform, beat frequency and fluid flow. The independent human sperm assertions remain separate. The existing NAS complex-assembly assertion is retained: PMID:26387594(https://pubmed.ncbi.nlm.nih.gov/26387594/) observes loss of axonemal DRC3 and infers improper assembly; its Results section on DRC3/LRRC48 localization and Discussion distinguish that inference from a reconstitution experiment. Complex assembly is now represented within the same integrated scaffold core.
The two NEW assertions rely on structural participation, not just knockout necessity. Detailed DRC1/DRC2/DRC4 architecture in PMID:37258679(https://pubmed.ncbi.nlm.nih.gov/37258679/) is algal; the human respiratory reconstruction places homologous models in density and lists O95995 in Supplementary Table 3 (physical PDF page 11). Human DRC3 loss and reduced bending amplitude corroborate a structural regulatory role. Current typed is_a/part_of closures show no redundant existing ancestor or descendant for either proposed term. The regulates edge from regulation of cilium movement to cilium movement is not an is_a edge. Same-role comparators were human DRC2/Q8IXS2 (GO:0003352 IBA and IMP) and algal DRC1/P0DL09 (other assembly annotations); their records are not treated as proof of an omission. The captured GO-CAM index has no target model imposing another role.
Ten verified HuRI interactions now use independently supported transcription-factor or coregulator partner classes. Inferred partner classes are identified as such. ZNF250-3 and ZNF792 retain their separately recorded fragment boundaries; matching parent-sequence checksums do not prove full-length reagents. SKIL’s corepressor mechanism is supported by the authentic abstract of PMID:10531062(https://pubmed.ncbi.nlm.nih.gov/10531062/), newly cached by the normal fetch command as abstract-only. This paper does not assay DRC4. ZNF629 remains a positive non-core pair because its current cis-regulatory binding assertion alone does not establish the stricter transcription-factor class. None of these refinements assigns transcriptional activity, native co-localization or a physiological regulatory role to DRC4.
PNKP binding is refined to kinase binding and phosphatase binding, consistent with DNA-end chemistry rather than protein phosphorylation. GSTA4, KDM1A, EHHADH and NDOR1 retain enzyme binding because the current complete descendant set has no fitting narrower class. CDC37’s client-contact evidence supports chaperone binding without requiring DRC4 to be its folding client; GOLGA2’s established microtubule-organizing activity supports the cytoskeletal regulatory class. The focused evidence audit records the literal pair counts, functional-class assertions, term-source hashes and public source URLs.
The two mouse-derived cytoplasm assertions remain non-core at their original resolution. Positive human HPA cytosolic staining is retained as a separate, nonexclusive observation, not relabeled as the mouse donor experiment. The source guide records the species, assay and access limits. Raw caches, earlier source extractions, previous notes and history remain unchanged.