EEF1A2

Summary: Conserved translation elongation factor activity.

Reason: The current PTHR23115 tree places the exact human Q05639 sequence at PTN002508371 below PTN000562490, which carries the literal GO:0003746 IBD. Its alignment exactly recovers all 463 residues. Human and ortholog experimental donor records support this conserved activity; donor count is not used to grade the phylogenetic judgement.

Summary: Combined electronic evidence supports elongation-factor activity.

Reason: The actual target matches InterPro IPR004539 across residues 1–444, and that entry maps to translation elongation factor activity. The second proximate source is the mouse Eef1a2/Ensembl ortholog, whose current record retains experimental elongation annotations. These are two traceable inputs to the combined annotation, not an unspecified ARBA rule.

Summary: Rabbit EEF1A2 supports conserved elongation-factor activity.

Reason: The literal donor Q71V39 is rabbit EEF1A2, with current direct experimental annotations to this activity and translational elongation. The inspected primary abstracts describe purified rabbit eEF1A2 and translation activity. The human target falls in the same EEF1A2 subfamily and retains the complete translation-factor architecture; the transfer does not imply identical nucleotide-exchange rates between paralogs.

Summary: GTP hydrolysis is part of the elongation cycle.

Reason: PTN000561878 is an actual ancestor of Q05639 in the current PTHR23115 tree and bears the literal GTPase IBD. The target preserves the G domain and nucleotide-binding motifs, and ortholog biochemical evidence supports the GTP-dependent elongation cycle. This is not a claim of an independent signaling GTPase.

Summary: The conserved GTP-binding domain supports GTPase activity.

Reason: InterPro IPR000795 is a verified target match at residues 5–242 and explicitly maps to GTPase activity. The complete human sequence, curated GTP-binding features and conserved elongation mechanism corroborate this mapping.

Summary: Ortholog biochemistry supports GTPase activity.

Reason: Rabbit Q71V39 is EEF1A2 and has direct experimental GTPase annotations. The inspected biochemical source abstracts distinguish nucleotide binding and exchange from translation activity, while the current donor record supplies the curated hydrolysis assertion. Human Q05639 retains the same GTPase architecture. The two dated raw GOA records seed one identical assertion; both original records remain preserved.

Summary: Contextual association with SARS-CoV nucleocapsid.

Reason: The original study reports human EF1α–SARS-CoV nucleocapsid association by two-hybrid, coimmunoprecipitation, far-Western and surface plasmon resonance. Its Methods distinguish the human fetal-liver cDNA constructs from endogenous EF1α purified from 293T cells; the latter is not independently paralog-resolved. The curated Q05639 interaction records include the 291–463 target fragment. This positive pathogen-protein association is retained as contextual binding. The assays do not establish a normal host exogenous-protein-binding function or an autonomous antiviral activity. IntAct EBI-25498109, EBI-25498133, EBI-25498269, EBI-25498163, EBI-25498187, EBI-25498249, EBI-25498294, EBI-25506724, EBI-25506746, EBI-25566262. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Binding to the serine/threonine protein kinase LRRK2.

Reason: The primary abstract reports LRRK2/EF1A coassociation and reduced LRRK2 kinase activity upon recombinant-protein co-incubation. The curated human Q05639–Q5S007 record specifies T7-tagged EEF1A2 and GST-tagged LRRK2 residues 970–2527 with mutation features; complete original Methods were not recovered. LRRK2 is independently established as a serine/threonine protein kinase. The positive pair supports the informative partner-class term without assigning kinase activity to EEF1A2 or extrapolating inhibition to every cellular setting. IntAct EBI-6509067. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Proposed replacements: protein serine/threonine kinase binding

Summary: Supported association with ABTB1.

Reason: The exact human EEF1A2–ABTB1 association is present in this source. The independent ABTB1 record describes probable regulatory roles but does not establish a specific enzyme, transporter, transcription-factor or other informative binding class for this decision. Retain the supported association as non-core; neither a functional regulatory effect nor an adaptor activity is inferred from the pair. IntAct EBI-10223683, EBI-10380096, EBI-10599180. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Association with thromboxane-A synthase TBXAS1.

Reason: The curator-supported IPI row is matched to the positive IntAct anti-tag coimmunoprecipitation record and the BioPlex 2.0 filtered network: TBXAS1 (P24557) is bait and EEF1A2 is prey, with model interaction score 0.841854892390855. No reciprocal edge for this pair appears in the saved directed network. Independent reviewed partner evidence establishes the class represented by GO:0019899; this refines the existing association without assigning the partner activity or its regulation to EEF1A2. BioPlex used technical duplicates and peptide, abundance and background-frequency filters, but nonspecific recovery of an abundant elongation factor and indirect association remain possible. The spectra, prey-discriminating peptides and exact bait insert were not independently reanalysed. BioPlex 3.0 includes reanalysis of 2.0, so the two sources are not independent confirmation. Retain the class refinement at the evidential resolution of the imported association, without a native or direct-interface claim.

Proposed replacements: enzyme binding

Summary: Supported association with TTLL12.

Reason: The exact source-specific pair is positive, but TTLL12 must not be classified as an enzyme solely from its tubulin-tyrosine-ligase-like family name. The current curated record explicitly notes absent tested tyrosine-ligase and histone-methyltransferase activities and an indirect role in tubulin modification. No informative molecular-function refinement is established by this pair. Retain the association as non-core, with the recorded binary-versus-complex scope preserved. IntAct EBI-21859130. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Association with the sodium/citrate transporter SLC13A5.

Reason: The curator-supported IPI row is matched to the positive IntAct anti-tag coimmunoprecipitation record and the BioPlex 2.0 filtered network: SLC13A5 (Q86YT5) is bait and EEF1A2 is prey, with model interaction score 0.877315845204938. No reciprocal edge for this pair appears in the saved directed network. Independent reviewed partner evidence establishes the class represented by GO:0044325; this refines the existing association without assigning the partner activity or its regulation to EEF1A2. BioPlex used technical duplicates and peptide, abundance and background-frequency filters, but nonspecific recovery of an abundant elongation factor and indirect association remain possible. The spectra, prey-discriminating peptides and exact bait insert were not independently reanalysed. BioPlex 3.0 includes reanalysis of 2.0, so the two sources are not independent confirmation. Retain the class refinement at the evidential resolution of the imported association, without a native or direct-interface claim.

Proposed replacements: transmembrane transporter binding

Summary: Supported association with ABTB1.

Reason: The exact human EEF1A2–ABTB1 association is present in this source. The independent ABTB1 record describes probable regulatory roles but does not establish a specific enzyme, transporter, transcription-factor or other informative binding class for this decision. Retain the supported association as non-core; neither a functional regulatory effect nor an adaptor activity is inferred from the pair. IntAct EBI-21859899. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Association with the sialyltransferase ST6GALNAC1.

Reason: The curator-supported IPI row is matched to the positive IntAct anti-tag coimmunoprecipitation record and the BioPlex 2.0 filtered network: ST6GALNAC1 (Q9NSC7) is bait and EEF1A2 is prey, with model interaction score 0.817084938392657. No reciprocal edge for this pair appears in the saved directed network. Independent reviewed partner evidence establishes the class represented by GO:0019899; this refines the existing association without assigning the partner activity or its regulation to EEF1A2. BioPlex used technical duplicates and peptide, abundance and background-frequency filters, but nonspecific recovery of an abundant elongation factor and indirect association remain possible. The spectra, prey-discriminating peptides and exact bait insert were not independently reanalysed. BioPlex 3.0 includes reanalysis of 2.0, so the two sources are not independent confirmation. Retain the class refinement at the evidential resolution of the imported association, without a native or direct-interface claim.

Proposed replacements: enzyme binding

Summary: Binding to the serine/threonine protein kinase LRRK2.

Reason: The original Methods and Results identify human Myc-tagged LRRK2 WT/G2019S affinity purification in human HEK293T cells, separately from the mouse astrocyte experiments. The exact EEF1A2 records are spoke-expanded physical associations, not purified binary contact. LRRK2 is independently a serine/threonine protein kinase, so the supported association can be refined to this partner class. The unavailable Table S1 workbook does not negate the explicit deposited pair records. IntAct EBI-21364596, EBI-21360986. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Proposed replacements: protein serine/threonine kinase binding

Summary: Supported association with ABTB1.

Reason: The exact human EEF1A2–ABTB1 association is present in this source. The independent ABTB1 record describes probable regulatory roles but does not establish a specific enzyme, transporter, transcription-factor or other informative binding class for this decision. Retain the supported association as non-core; neither a functional regulatory effect nor an adaptor activity is inferred from the pair. The deposited mutation feature is kept on its literal participant side: uniprotkb:Q05639 has -; uniprotkb:Q969K4 has mutation with no effect:386-386(rs73203006). IntAct EBI-22131807. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Supported association with TTLL12.

Reason: The exact source-specific pair is positive, but TTLL12 must not be classified as an enzyme solely from its tubulin-tyrosine-ligase-like family name. The current curated record explicitly notes absent tested tyrosine-ligase and histone-methyltransferase activities and an indirect role in tubulin modification. No informative molecular-function refinement is established by this pair. Retain the association as non-core, with the recorded binary-versus-complex scope preserved. IntAct EBI-25256689, EBI-22480223, EBI-22936623. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Supported association with ABTB1.

Reason: The exact human EEF1A2–ABTB1 association is present in this source. The independent ABTB1 record describes probable regulatory roles but does not establish a specific enzyme, transporter, transcription-factor or other informative binding class for this decision. Retain the supported association as non-core; neither a functional regulatory effect nor an adaptor activity is inferred from the pair. IntAct EBI-24706037, EBI-25294003, EBI-24042006, EBI-23749529. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Contextual binding to an HTT amino-terminal construct.

Reason: The positive curated two-hybrid records use an HTT 1–546 construct and record a polyglutamine-region mutation that changes interaction strength. Reference-sequence identity is authenticated, but this does not make the tested fragment a full-length HTT assay. The evidence does not establish a specific binding class or protein-aggregation function of EEF1A2; retain the source assertion as non-core. IntAct EBI-25959972, EBI-26114293, EBI-26268424. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Association with thromboxane-A synthase TBXAS1.

Reason: The curator-supported IPI row is matched to the positive IntAct anti-tag coimmunoprecipitation record and the BioPlex 3.0 filtered network: TBXAS1 (P24557) is bait and EEF1A2 is prey, with model interaction score 0.892770103025646. No reciprocal edge for this pair appears in the saved directed network. Independent reviewed partner evidence establishes the class represented by GO:0019899; this refines the existing association without assigning the partner activity or its regulation to EEF1A2. BioPlex used technical duplicates and peptide, abundance and background-frequency filters, but nonspecific recovery of an abundant elongation factor and indirect association remain possible. The spectra, prey-discriminating peptides and exact bait insert were not independently reanalysed. BioPlex 3.0 includes reanalysis of 2.0, so the two sources are not independent confirmation. Retain the class refinement at the evidential resolution of the imported association, without a native or direct-interface claim.

Proposed replacements: enzyme binding

Summary: Supported association with TTLL12.

Reason: The exact source-specific pair is positive, but TTLL12 must not be classified as an enzyme solely from its tubulin-tyrosine-ligase-like family name. The current curated record explicitly notes absent tested tyrosine-ligase and histone-methyltransferase activities and an indirect role in tubulin modification. No informative molecular-function refinement is established by this pair. Retain the association as non-core, with the recorded binary-versus-complex scope preserved. IntAct EBI-54762512, EBI-55016128. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Association with the sodium/citrate transporter SLC13A5.

Reason: The curator-supported IPI row is matched to the positive IntAct anti-tag coimmunoprecipitation record and the BioPlex 3.0 filtered network: SLC13A5 (Q86YT5) is bait and EEF1A2 is prey, with model interaction score 0.947898505501937. No reciprocal edge for this pair appears in the saved directed network. Independent reviewed partner evidence establishes the class represented by GO:0044325; this refines the existing association without assigning the partner activity or its regulation to EEF1A2. BioPlex used technical duplicates and peptide, abundance and background-frequency filters, but nonspecific recovery of an abundant elongation factor and indirect association remain possible. The spectra, prey-discriminating peptides and exact bait insert were not independently reanalysed. BioPlex 3.0 includes reanalysis of 2.0, so the two sources are not independent confirmation. Retain the class refinement at the evidential resolution of the imported association, without a native or direct-interface claim.

Proposed replacements: transmembrane transporter binding

Summary: Supported association with ABTB1.

Reason: The exact human EEF1A2–ABTB1 association is present in this source. The independent ABTB1 record describes probable regulatory roles but does not establish a specific enzyme, transporter, transcription-factor or other informative binding class for this decision. Retain the supported association as non-core; neither a functional regulatory effect nor an adaptor activity is inferred from the pair. IntAct EBI-54763155, EBI-54859585. Reference-sequence ROGIDs match the official participant sequences; this is not an experimental-insert sequencing claim.

Summary: Association with the sialyltransferase ST6GALNAC1.

Reason: The curator-supported IPI row is matched to the positive IntAct anti-tag coimmunoprecipitation record and the BioPlex 3.0 filtered network: ST6GALNAC1 (Q9NSC7) is bait and EEF1A2 is prey, with model interaction score 0.79450338033373. No reciprocal edge for this pair appears in the saved directed network. Independent reviewed partner evidence establishes the class represented by GO:0019899; this refines the existing association without assigning the partner activity or its regulation to EEF1A2. BioPlex used technical duplicates and peptide, abundance and background-frequency filters, but nonspecific recovery of an abundant elongation factor and indirect association remain possible. The spectra, prey-discriminating peptides and exact bait insert were not independently reanalysed. BioPlex 3.0 includes reanalysis of 2.0, so the two sources are not independent confirmation. Retain the class refinement at the evidential resolution of the imported association, without a native or direct-interface claim.

Proposed replacements: enzyme binding

Summary: Supported association with TTLL12.

Reason: The exact source-specific pair is positive, but TTLL12 must not be classified as an enzyme solely from its tubulin-tyrosine-ligase-like family name. The current curated record explicitly notes absent tested tyrosine-ligase and histone-methyltransferase activities and an indirect role in tubulin modification. No informative molecular-function refinement is established by this pair. Retain the association as non-core, with the recorded binary-versus-complex scope preserved. The original AP-MS network contains the exact EEF1A2–TTLL12 edge. The AP-MS node symbols and Ensembl identifiers identify this pair; no ROGID or experimental-insert sequence is supplied by that graph.

Summary: Supported association with ABTB1.

Reason: The exact human EEF1A2–ABTB1 association is present in this source. The independent ABTB1 record describes probable regulatory roles but does not establish a specific enzyme, transporter, transcription-factor or other informative binding class for this decision. Retain the supported association as non-core; neither a functional regulatory effect nor an adaptor activity is inferred from the pair. The original AP-MS network contains the exact EEF1A2–ABTB1 edge. The AP-MS node symbols and Ensembl identifiers identify this pair; no ROGID or experimental-insert sequence is supplied by that graph.

Summary: GTP binding supports the elongation-factor cycle.

Reason: The three literal InterPro sources are verified target matches and carry the GTP-binding mapping. The human record also contains experimentally structured or conserved nucleotide-contact features. GTP binding is part of the same translation mechanism as hydrolysis and aminoacyl-tRNA delivery, not a separate signaling role.

Summary: Nuclear exclusion is supported in the reported localization assay.

Reason: This is a negated annotation. The original Methods explicitly name human eEF1A2/S2 cDNA, and Results/Figure 2 describe GFP-tagged S2 in BHK cells with cytoplasmic localization and strong nuclear exclusion. The endogenous antibody experiments are broader EF1α measurements and are kept separate. Retain the assay-qualified absence; it does not assert impossibility of nuclear entry under every perturbation or cell state.

Summary: Human eEF1A2/S2 is cytoplasmic in the source assay.

Reason: The actual original Methods identify human eEF1A2/S2, and the GFP-tagged S2 localization experiment in BHK cells supports cytoplasm with nuclear exclusion. This experiment does not by itself discriminate soluble cytosol from peripheral membrane-associated pools. Keep the source-specific broad compartment instead of forcing a narrower location from unrelated assays.

Summary: Cytoplasmic localization transfers coherently from mouse Eef1a2.

Reason: The literal mouse donor P62631 and Ensembl protein identify Eef1a2. Its current record retains an independent mouse neuronal cytoplasm IDA assertion in addition to propagated rows. The direct human S2 localization study independently agrees. Reciprocal propagated records are recorded but are not treated as the only evidence or as a reason to reject the transfer.

Summary: ER-associated translation supports the source localization.

Reason: The original study examines human HEK293F ER-associated translation and separately prepared ribosomal particles. Deposited model 8B6Z entity 1 specifies the exact human Q05639 sequence; the rabbit starting model 4C0S is not a wrong-paralog explanation. Native-particle discrimination between close eEF1A paralogs is limited. Retain peripheral, ribosome-associated ER membrane localization without claiming an integral membrane protein.

Summary: The UniProt ER-membrane mapping retains its experimental source.

Reason: The literal SL-0097 mapping is traceable to the current UniProt ER-membrane comment with experimental evidence from PMID36697828. Its meaning here is translation-associated peripheral localization, consistent with the deposited human EEF1A2 model, not membrane insertion or exclusivity.

Summary: Conserved participation in the eukaryotic elongation-factor complex.

Reason: The current mouse Eef1a2 donor record explicitly retains a curator-assigned IDA for this complex, and the term definition includes the alpha elongation-factor component with the exchange-factor subunits. The historical biochemical abstract concerns purified rabbit eEF1A2, so it is not described as a newly verified mouse complex-reconstitution assay. Retain the conserved complex assertion without making stable or obligatory association a claim of the core model.

Summary: Refine translation to the elongation step performed by EEF1A2.

Reason: The literal translation IBD at PTN000561878 is authentic and the target descends from it. EEF1A2 lies within the more specific EF1A clade PTN000562490, with an independently placed translational-elongation IBD and a sequence-authenticated elongation-factor architecture. The broad source assertion is sound; the existing elongation term identifies the step performed by this target rather than initiation or termination.

Proposed replacements: translational elongation

Summary: Conserved translational elongation.

Reason: The current tree and MSA place exact Q05639 below PTN000562490 and the literal IBD at that node specifies translational elongation. Current experimental donor records and biochemical source abstracts agree. There is no target-specific loss indicated by the full 463-residue alignment.

Summary: InterPro maps the verified EF1A family to elongation.

Reason: IPR004539 is an actual target family match and its current GO mapping includes translational elongation. This agrees with aminoacyl-tRNA delivery rather than merely an indirect requirement for protein synthesis.

Summary: Rabbit EEF1A2 biochemistry supports elongation.

Reason: Q71V39 is verified rabbit EEF1A2 and carries direct experimental elongation assertions. The biochemical abstracts describe purified eEF1A2 participating in an in vitro translation system. The homologous human sequence retains the complete architecture, supporting the source transfer without equating all kinetic parameters of the two eEF1A paralogs.

Summary: The review explicitly includes EEF1A2 in translational elongation.

Reason: The cited article identifies both mammalian eEF1A variants and states that they have similar translation activity. It is a review, as its abstract explicitly says, despite the normal cache metadata classifying it as primary research. Retain this NAS as a supporting statement for the experimentally grounded elongation function; it is not counted as a new direct assay.

Summary: Use the specific translation elongation factor activity.

Reason: The accessible abstract of the chromosome-mapping paper explicitly states that both human genes can function in translation; it does not provide a separately inspected RNA-binding assay. The existing broad translation-factor assertion is consistent with the independent EF1A phylogeny and biochemistry. Refine it to the established elongation-factor activity, without claiming that this 1996 paper directly measured the complete mechanism.

Proposed replacements: translation elongation factor activity

Summary: EEF1A2 directly activates the lipid kinase PI4KB.

Reason: The source partner is exactly PI4KB/Q9UBF8, a phosphatidylinositol lipid kinase, whereas the imported protein-kinase-binding definition requires a protein substrate. The preserved PMID:17088255 Abstract, publications/PMID_17088255.md lines 41–44, reports direct EEF1A2 binding to PI4KB and increased PI4KB activity with purified recombinant EEF1A2. These accessible observations support GO:0098744. Separately read original Methods and Results provide the GST-removal, albumin-control and endogenous MCF7 details, located in EEF1A2-source-evidence.yaml; those details are not attributed to the abstract cache. This remains an activator function, not autonomous lipid-kinase catalysis by EEF1A2.

Proposed replacements: 1-phosphatidylinositol 4-kinase activator activity

Summary: Synaptic localization is a supported neuronal context.

Reason: The exact mouse Eef1a2 donor currently carries SynGO EXP/IDA synapse annotations from a postsynaptic-density proteomic study. The abstract establishes the proteomic sample context, while the source-specific peptide table was not independently reconstructed. Retain the conservative human ortholog localization as a contextual pool rather than converting it into a distinct synaptic signaling mechanism or claiming direct human synapse imaging.

Summary: Direct PI4KB activation supports positive lipid-kinase regulation.

Reason: The preserved PMID:17088255 Abstract, publications/PMID_17088255.md lines 41–44, directly reports increased PI4KB activity with purified recombinant EEF1A2; lines 45–48 add the cellular expression and depletion results. Separately read original Methods and Results resolve the purified human pair and endogenous MCF7 experiments, with locators in EEF1A2-source-evidence.yaml. The positive sign is measured rather than inferred solely from necessity. PI4KB performs lipid phosphorylation; EEF1A2 is its activator, and physiological importance in normal tissues remains to be resolved.

Summary: EF1α lysosomal topology is positive, but EEF1A2 identity is unresolved.

Reason: The original Results show an EF1α pool on the cytoplasmic side of lysosomal membranes using fractionation, extraction and protease susceptibility. These experiments used rat liver/hepatocytes and mouse fibroblasts. The Discussion explicitly leaves unresolved which of the two EF1α gene products or modified forms supplies this pool. Preserve the imported human EEF1A2 assertion and genuine topology result, but the recovered evidence does not resolve its paralog-specific attribution. The rat-liver and fibroblast context raises a paralog-attribution question, but tissue expression patterns alone do not identify the measured EF1α protein or establish that EEF1A2 is absent.

Summary: The EF1α-dependent CMA mechanism lacks resolved EEF1A2 attribution.

Reason: The source reports EF1α/GFAP association and GTP-dependent exchange at lysosomes that modulates the LAMP-2A translocation machinery. The authors could not interpret EF1α knockdown because of broader lysosomal alterations and instead used blocking-antibody experiments. In rat and mouse systems, the paper explicitly leaves the responsible EF1α gene product unresolved. This is a real family-level mechanism, not evidence that EEF1A2 lacks it; retain uncertainty on the imported human EEF1A2-specific NAS. The rat-liver and fibroblast context raises a paralog-attribution question, but tissue expression patterns alone do not identify the measured EF1α protein or establish that EEF1A2 is absent.