eryCV

Summary: Root-level "catalytic activity" - uninformative; EryCV is a specific iron-sulfur/SAM enzyme.

Reason: Under the project rule GO:0003824 is KEEP_AS_NON_CORE where no reaction is known, not MARK_AS_OVER_ANNOTATED: the row is true and is as precise as the evidence allows. EryCV is Predicted (UniProt PE 4), its only name is an unreviewed submitter annotation, and the ammonia-lyase row in this review is UNDECIDED, so there is no specific reaction term to refine to.

Summary: Ammonia-lyase activity from an InterPro signature. This conflicts with the enzyme's name (3,4-enoyl reductase) and its radical-SAM (Fe-S/SAM) cofactor profile; DesII-type radical-SAM enzymes can catalyze dehydration/deamination, but the precise reaction here is unverified.

Reason: Conflicting electronic assignments (ammonia-lyase vs 3,4-enoyl reductase) with no accessible experimental characterization; the radical-SAM cofactor binding is, however, well supported.

Summary: Involvement in macrolide deoxysugar biosynthesis, at the level family membership supports: the record matches the InterPro DesII entry this row fired on, plus two member-database signatures that name the same family and one that names the substrate, and DesII-family radical SAM enzymes act on the TDP-deoxysugar intermediates of macrolide amino-sugar (desosamine) pathways. True and as precise as the evidence allows, but not a core function.

Reason: Kept, and grounded on family membership rather than on the gene name, but non-core. The action follows the reason: this review asserts no core molecular function and no core process, so a pathway term with no established activity behind it cannot be the core function the ACCEPT it previously carried would claim. CLAUDE.md puts the disjunction directly -- where a term is already as precise as the evidence allows, keep it, ACCEPT if core and otherwise KEEP_AS_NON_CORE -- and GO:0003824 in this same review is kept non-core on exactly this argument, as are every retained row in PSEPK/PP_1969 and PSEPK/PP_2482, the members eryCV now matches. On the evidence: GO:0033068 is an InterPro2GO mapping (GO_REF:0000002) whose WITH/FROM names InterPro:IPR016863, so the first quote below is not merely consistent with the mapping, it is the entry the mapping fired on. Three member-database signatures agree, two of them naming the family (NCBIfam TIGR04426 rSAM_desII, SFLD SFLDG01116 DesII-like) and one naming the substrate itself (SFLD SFLDF00425 dTDP-4-amino-4,6-dideoxy-D-glucose, the desosamine precursor). The repository policy treats family and domain classification as good support for membership, and membership as an argument for function, so these place the protein in macrolide deoxysugar biosynthesis independently of the unreviewed submitter gene name. Note what this does and does not support: the genus term GO:0033068 follows from the family, while the erythromycin-specific child GO:1901115 would need this particular cluster to be the erythromycin one, which rests only on that submitter name -- which is why GO:1901115 is neither asserted nor proposed. The reaction itself remains unestablished, so this row says where the protein acts, not what it does.

Summary: Iron-sulfur cluster binding. This row's InterPro2GO mapping fired on IPR007197 rSAM, named in the GOA row's WITH/FROM column, and the record matches that entry alongside the Pfam PF04055 Radical_SAM signature. Correct but general: the radical SAM cluster is [4Fe-4S].

Reason: Following the RADICAL_SAM project rule for the generic domain-derived rows (projects/RADICAL_SAM.md): correct but general. The entry has no COFACTOR block, but the radical SAM signatures it matches define the [4Fe-4S] cluster that reductively cleaves SAM, so GO:0051539, already carried by this entry, is the descendant the family always satisfies.

Proposed replacements: 4 iron, 4 sulfur cluster binding

Summary: [4Fe-4S] cluster binding, the hallmark cofactor of a radical SAM enzyme, inferred from signature matches rather than measured. This row's InterPro2GO mapping fired on the InterPro entry IPR016863 DesII, named in the GOA row's WITH/FROM column; the record matches that entry alongside the NCBIfam TIGR04426 rSAM_desII signature, and the Pfam PF04055 Radical_SAM match corroborates the cluster independently. Those three are what this row quotes. Which member-database signature InterPro integrates into which entry is not recorded in this file, so no such pairing is asserted.

Reason: The project rule ACCEPTs GO:0051539 for active members and keeps it non-core where catalysis is not established. Nothing is established for EryCV: the protein is Predicted (PE 4), no reaction has been assayed, and the cluster itself is inferred from family signatures rather than measured. Treated as the project treats the other uncharacterised members (PSEPK/PP_1969, PSEPK/PP_2482, TYW1B, aslB, ydeM), which assert no core molecular function.