COA4

Summary: Phylogenetic propagation of the complex IV assembly role within the narrow COA4 family, seeded in part from this very gene's own experimental annotations (SGD:S000004208).

Reason: Sound node placement, and the target carries direct IMP and IGI evidence for the same term. The appearance of SGD:S000004208 in the WITH/FROM is expected and correct: the yeast experimental annotations are among the descendant evidences the PAINT curator used to place the IBD, and the IBA then adds the further claim that the function is inherited rather than lineage-specific. This is the core biological process.

Summary: Phylogenetic propagation of intermembrane space localization, matching the direct experimental localization of this protein.

Reason: Corroborated on the target itself by EXP, two IDA rows and TAS. The is_active_in qualifier is right: Coa4 acts in the IMS, contacting the IMS-facing domain of Cox11.

Summary: Automated inner-membrane localization from UniProt subcellular location. Coa4 is peripherally associated with the inner membrane rather than integral.

Reason: Consistent with UniProt and with the reported inner-membrane association, but in slight tension with the Bax-release IMS proteomics that recovers Coa4 as a releasable intermembrane space protein. Best read as peripheral association; the IMS term is the primary and more informative compartment.

Summary: InterPro-based localization from the Coa4-like family signature IPR039870, consistent with the direct experimental evidence.

Reason: The Coa4-like family is narrow and functionally homogeneous, so the InterPro2GO mapping is reliable here, and the target has independent EXP and IDA support for the same term.

Summary: InterPro-based process assignment from IPR039870, duplicating the IBA and IMP calls.

Reason: Correct and redundant rather than erroneous. The family signature genuinely predicts this process for this family, and the target has direct IMP evidence.

Summary: Direct experimental placement in the intermembrane space by Bax-release proteomic profiling of isolated mitochondria.

Reason: A targeted IMS proteomics method with high coverage of the known soluble IMS proteome, and the strongest single piece of localization evidence for this protein. Coa4 is not named in the cached body text - the assignment is in the study's protein tables - so the curator call is accepted rather than second-guessed.

Summary: Nuclear localization from the genome-wide C-terminal GFP fusion library, contradicted by every other localization source for this protein.

Reason: Flagged because it conflicts with everything else known about this protein, not because the evidence behind it could not be inspected. COA4 is a twin CX9C substrate of the MIA40-ERV1 disulfide relay, and there is no described route by which such a protein reaches the nucleus: import commits it to the intermembrane space, where its two disulfides are oxidatively trapped. The call is contradicted by Bax-release IMS proteomics (EXP), direct assay localization (IDA), the phylogenetic and InterPro assignments, Reactome, and the UniProt subcellular location, and no other source of any kind places this protein in the nucleus. The most likely explanation is that C-terminal GFP fusion blocked import and the mislocalized fusion was scored outside mitochondria - the same library gives this protein a cytoplasmic call too. Marked over-annotated rather than removed so a curator with access to the original images can make the final call; REMOVE would be defensible. No supporting_text is attached: the cached record is abstract-only and does not name this protein, so no verbatim quotation can carry the claim - the paper's contribution here is the genome-wide localization call itself, which lives in its image dataset.

Summary: Cytoplasmic localization from the genome-wide C-terminal GFP fusion library. A cytosolic pool is expected for a MIA40 substrate, which dwells in the cytosol before import.

Reason: Mia40-pathway substrates genuinely occupy the cytosol before import - twin CX9C proteins remain there for several minutes while awaiting the disulfide relay - so a cytosolic signal for this protein is biologically expected rather than anomalous. The C-terminal GFP tag used by this library can also block import, so this row may partly report a tagging artifact, but nothing known about the protein contradicts a genuine cytosolic pool. Accepted as a correct compartment and kept non-core because the functionally relevant pool - the one that acts on the copper delivery pathway - is the intermembrane space one. No supporting_text is attached: the cached record is abstract-only and does not name this protein, so no verbatim quotation can carry the claim - the paper's contribution here is the genome-wide localization call itself, which lives in its image dataset.

Summary: Reactome traceable-author-statement placing Coa4 in the intermembrane space, agreeing with the direct experimental evidence.

Reason: Adds no independent evidence but is consistent with the EXP and IDA rows. Harmless and correct.

Summary: SGD's explicit 'no data' placeholder for molecular function.

Reason: Correct and worth preserving. No catalytic or binding activity has ever been demonstrated for Coa4, and metallochaperone activity is positively excluded because it lacks the copper-coordinating cysteine motif of Cox17. This ND row is an accurate statement of the state of knowledge, and it independently corroborates the decision in the human COA4 review not to assert a molecular function term. It should be retired only when a real molecular function is demonstrated, not replaced with a generic binding term. No supporting_text is attached: this row asserts the absence of demonstrated activity, and no quotation can evidence an absence. The argument for keeping it is the reasoning above, not a citation.

Summary: Direct assay localization to the intermembrane space from the systematic twin Cx9C family survey, which also established Mia40-dependent import for this protein.

Reason: Consistent with all other localization evidence, and mechanistically grounded: the twin Cx9C motif and Mia40 dependence are the signature of this compartment.

Summary: Direct assay localization to the intermembrane space in the study that characterized and renamed this protein.

Reason: The primary characterization of Coa4, which explicitly reports intermembrane space localization with inner membrane association.

Summary: Evidence for a role in respiratory chain complex assembly at the general level; the gene's complex IV-specific role is annotated separately.

Reason: Faithful to the cited study, which assayed respiratory chain function generally rather than complex IV specifically, so this is a correct annotation and not a defect. The gene also carries the complex IV-specific GO:0033617 from focused studies, so nothing is lost by leaving this parent term in place. Treated the same way as the redundant-parent GO:0008535 row on COX19: a correct-but-general parent is kept non-core, not rewritten. Re-pointing it would be a curation-policy change rather than a correction.

Summary: Mutant phenotype evidence that cells lacking Coa4 have depressed cytochrome c oxidase activity, with normal Cox1 maturation and a normal Shy1-stabilized Cox1 assembly intermediate.

Reason: The primary experimental basis for the core annotation. It also localizes Coa4's action downstream of Cox1 synthesis and hemylation, which is what later placed it at the copper insertion step.

Summary: Genetic interaction evidence: COA4 was isolated as an allele-specific suppressor of the Shy1 G137E mutant, and high-copy CYC1 restores respiratory function to coa4-null cells.

Reason: Two genuine genetic interactions (SHY1, SGD:S000003344; CYC1, SGD:S000003809) from a focused suppressor screen, supporting a role in complex IV assembly. The CYC1 interaction remains mechanistically unexplained and was called a major clue by the authors; it has not been followed up since.

Summary: Mutant phenotype evidence that deletion of COA4 causes cytochrome c oxidase assembly defects, specifically loss of Cox1 and rapid degradation of Cox2 and Cox3.

Reason: Independent IMP confirmation of the core annotation in a different experimental setting. Note the study's own caveat, which matters for reading any later rescue experiment: reductants restore respiratory growth of the mutant without restoring complex IV assembly, so growth rescue alone is not evidence of restored assembly.

Summary: Genetic interaction with CMC1: both twin CX9C mutants were recovered together as reductant-suppressed respiratory mutants with cytochrome c oxidase assembly defects.

Reason: A real shared-phenotype genetic interaction (CMC1, SGD:S000001620) supporting the complex IV assembly role. It reflects convergent phenotype rather than a demonstrated pathway relationship between the two proteins.