lipA

Summary: Generic catalytic activity is true but uninformative for LipA.

Reason: Following the RADICAL_SAM project rule for the generic domain-derived rows (projects/RADICAL_SAM.md): the row is true, not an over-annotation, so it is refined rather than flagged. LipA has a defined reaction, lipoate synthase activity (EC 2.8.1.8), which this entry already carries.

Proposed replacements: lipoate synthase activity

Summary: Cytoplasmic localization is plausible context for bacterial LipA.

Reason: UniProt places the protein in the cytoplasm, but localization is secondary to its radical-SAM lipoate synthase activity.

Summary: LipA participates directly in endogenous lipoate biosynthesis.

Reason: LipA is the sulfur-insertion reaction in the endogenous two-step route that produces protein-bound lipoyl groups. GO:0009107 was obsoleted by GO (2026-08-22, replaced_by GO:0009249 protein lipoylation, obsoleted because term usage was inconsistent), so the annotation should move to the replacement term. GOA already carries an independent GO:0009249 row for this gene, so the replacement merges onto an existing annotation rather than asserting a new one.

Proposed replacements: protein lipoylation

Summary: Protein lipoylation is the immediate biological process for LipA.

Reason: The enzyme directly converts octanoylated lipoyl-domain lysines into lipoylated derivatives.

Summary: Sulfurtransferase activity is a correct but redundant parent term.

Reason: Following the RADICAL_SAM project rule for the generic domain-derived rows (projects/RADICAL_SAM.md): sulfurtransferase activity is the right chemistry pitched without the substrate, and it is not false, so MODIFY rather than MARK_AS_OVER_ANNOTATED. Replaced by GO:0016992, as in human LIAS and BACSU/lipA.

Proposed replacements: lipoate synthase activity

Summary: Lipoate synthase activity is the specific LipA catalytic function.

Reason: UniProt records the EC 2.8.1.8/Rhea reaction that converts octanoyl-lysyl protein to dihydrolipoyl-lysyl protein.

Summary: Iron-sulfur cluster binding is correct but less precise than the 4Fe-4S term.

Reason: Following the RADICAL_SAM project rule for the generic domain-derived rows (projects/RADICAL_SAM.md): correct but general, and never MARK_AS_OVER_ANNOTATED, since the protein does bind iron-sulfur clusters. Both are [4Fe-4S], so GO:0051539, already carried here, is the right descendant.

Proposed replacements: 4 iron, 4 sulfur cluster binding

Summary: Binding two 4Fe-4S clusters is required mechanistic context for LipA.

Reason: Core cofactor binding under the project rule: the radical SAM [4Fe-4S] cluster is what generates the 5'-deoxyadenosyl radical and the second [4Fe-4S] cluster is the sulfur donated to the product, so neither is ancillary to lipoate synthase activity. Catalysis is established for this protein by the HAMAP family rule and the E. coli and human orthologues, so the row is ACCEPT, not KEEP_AS_NON_CORE.

Summary: lipA binds S-adenosyl-L-methionine at the unique iron of its radical SAM [4Fe-4S] cluster and cleaves it to the 5'-deoxyadenosyl radical; two SAM are consumed per lipoyl group formed.

Reason: Not in GOA. Added under the RADICAL_SAM treatment rule, which treats GO:1904047 as a core cosubstrate-binding function of every radical SAM enzyme whose catalysis is established. This is a molecular function the protein itself performs, so the participation test is met directly: lipA binds SAM and cleaves it, rather than merely being required for a process that consumes it. UniProt's curated cofactor note records the exchangeable S-adenosyl-L-methionine on the radical SAM cluster and the entry's binding features name the 4Fe-4S-S-AdoMet ligand, so the claim rests on curated record content rather than on an electronic GO line. Coded IC because the inference is the reviewer's, drawn from that curated content and the established lipoate synthase activity rather than from a SAM-binding assay on this protein. Comparator check, against GOA rather than against this project's own proposals: GOA does annotate GO:1904047 to radical SAM enzymes in exactly this role, by UniRule -- ECOLI/queE (UR000100349), ECOLI/moaA and PSEPK/moaA (UR000101305), METEA/pqqE and PSEPK/pqqE (UR000098303). So the absence here is patchy UniRule coverage, which rules happen to carry the term, not a GO convention against annotating a radical SAM enzyme's own cosubstrate. That is also why those five genes have an existing row to ACCEPT or MODIFY while this one needs a NEW row.