No deep-research provider was available; research done by hand. PubMed
E-utilities returned HTTP 500 on 2026-11-05, so the search was limited to the
cached publications (grep of publications/ for EftM/PA4178).
| PTN node | node term | n_seeds | leaf IBAs | seed annotation (term, evidence, PMID) | verdict | rationale | other seeds | IBD stands without this seed? |
|---|---|---|---|---|---|---|---|---|
| PTN000010968 | GO:0008168 methyltransferase activity | 10 | 412 (77 taxa) | GO:0008168, IDA, PMID:26677219 (PseudoCAP) | SOUND | Full text read. Purified EftM methylates EF-Tu with SAM; MS maps trimethyl-Lys5. The seed is under-specific (the paper supports GO:0016279), but a descendant is valid IBD evidence for the generic node term. | E. coli UbiE P0A887, human COQ5 Q5HYK3, yeast COQ5 S000004578, Synechocystis MPBQ MTase P74388, A. nidulans LaeA C8VQG9, Candida CAL0000191608, three Arabidopsis loci | Yes. Several seeds are well-characterised methyltransferases; generic methyltransferase activity suits a broad class I MTase node. No GO_CENTRAL defect for this node. |
Separate problem (not a PseudoCAP seed): PA4178 also receives UbiE/COQ5 IBAs
(GO:0008425, GO:0006744) from the child node PTN001297884. These are wrong for
EftM, a protein lysine methyltransferase. Both rows are REMOVE, with one
GO_CENTRAL defect (PROBABLE) recorded on the GO:0006744 row.