# CAZy family -> GO molecular-function mapping (SSSOM) -- SEED # # Forward-propagation source for the Glycobiology project (see ../GLYCOBIOLOGY.md and # GLYCOBIOLOGY-resource-reuse.md). Maps a CAZy carbohydrate-active-enzyme FAMILY to the GO # molecular-function term that best describes its catalytic activity -- the glyco analogue of # interpro2go / ec2go. CAZy reaches GO only indirectly today (via UniProt/EC/InterPro); there is no # official cazy2go, so this curated set IS the mapping. # # This is a SEED: it covers only the five glycosyltransferase families that the project's exemplar # genes belong to (each row backed by the corresponding completed gene review). Scaling to the full # ~190 GT/GH/PL/CE families is the follow-up (seedable from the dbCAN-sub subfamily->EC->substrate # table, then EC->ec2go). # # Predicate semantics (mirrors projects/INTERPRO/interpro2go.sssom.yaml): # * skos:exactMatch -- the GO term holds for ~all (animal) members of the family; family is # mono-specific at this altitude. Ready-to-use family->GO MF. # * skos:narrowMatch -- the GO term is NARROWER than the family: it describes the dominant # subfamily but NOT all family members, so the family is too coarse and # SUBFAMILY resolution (dbCAN-sub / EC) is needed before propagating. # # Redundancy filter (the "trivial" part): a row is REDUNDANT if the family's EC already maps to the # same GO term via ec2go (GO_REF:0000003) or an InterPro entry for the family maps to it via # interpro2go (GO_REF:0000002). Redundant rows add no marginal annotation; the genuine contribution # is rows whose GO term is NOT reachable that way. Redundancy status is noted per row in `comment` # (qualitative here; compute live against ec2go closure to confirm, as projects/RHEA does for EC). # # Provenance: every CAZy family + EC is from the CAZy/CAZypedia family page; every GO id/label is # QuickGO-verified non-obsolete; every activity assignment is backed by the cited completed gene # review in genes/human//. curie_map: CAZy: http://www.cazy.org/ GO: http://purl.obolibrary.org/obo/GO_ skos: http://www.w3.org/2004/02/skos/core# semapv: https://w3id.org/semapv/vocab/ sssom: https://w3id.org/sssom/ mapping_set_id: https://w3id.org/ai4curation/ai-gene-review/mappings/cazy2go-seed mapping_set_title: CAZy family -> GO molecular function (Glycobiology seed) mapping_set_description: >- Curated CAZy-family -> GO molecular-function mappings seeded from the Glycobiology project's exemplar glycosyltransferases. exactMatch rows are mono-specific families ready to propagate; narrowMatch rows flag families that are too coarse (the GO term applies only to a subfamily, so subfamily/EC resolution is required). Intended as a forward-propagation source analogous to interpro2go, with redundancy against ec2go/interpro2go to be closure-filtered before use. license: https://creativecommons.org/licenses/by/4.0/ creator_label: - AI Gene Review project mappings: # ===== exactMatch: mono-specific families, ready-to-use family->GO MF ===== - subject_id: CAZy:GT13 subject_label: "Glycosyltransferase Family 13 (N-acetylglucosaminyltransferase I, GnT-I)" predicate_id: skos:exactMatch predicate_label: exact match object_id: GO:0003827 object_label: alpha-1,3-mannosylglycoprotein 2-beta-N-acetylglucosaminyltransferase activity mapping_justification: semapv:ManualMappingCuration comment: >- GT13 is mono-specific (GnT-I, EC 2.4.1.101) -- the committed step of complex N-glycan synthesis. Backed by the MGAT1 (P26572) review (core MF = GO:0003827, exactMatch). LIKELY REDUNDANT with ec2go (EC 2.4.1.101 -> GO:0003827 is an established ec2go row); marginal value only where an entry lacks the EC. - subject_id: CAZy:GT65 subject_label: "Glycosyltransferase Family 65 (protein O-fucosyltransferase)" predicate_id: skos:exactMatch predicate_label: exact match object_id: GO:0046922 object_label: peptide-O-fucosyltransferase activity mapping_justification: semapv:ManualMappingCuration comment: >- In animals GT65 = POFUT1 (GDP-fucose protein O-fucosyltransferase 1, EC 2.4.1.221), which adds O-fucose to folded EGF repeats in the ER. Backed by the POFUT1 (Q9H488) review. exactMatch for metazoa; note GT65 also contains a kinetoplastid alpha-1,2-fucosyltransferase, so restrict to the animal POFUT1 subfamily when propagating outside metazoa. - subject_id: CAZy:GT29 subject_label: "Glycosyltransferase Family 29 (sialyltransferases)" predicate_id: skos:exactMatch predicate_label: exact match object_id: GO:0008373 object_label: sialyltransferase activity mapping_justification: semapv:ManualMappingCuration comment: >- GT29 is the sialyltransferase family (ST3/ST6/ST8 Gal/GalNAc subfamilies, EC 2.4.3.x). The family scope == the GO class 'sialyltransferase activity', so exactMatch at the CLASS level -- but the specific linkage child (e.g. GO:0003835 alpha-2,6 for ST6GAL1) needs subfamily/EC resolution. Backed by the ST6GAL1 (P15907) review. The class term is partly reachable via ec2go from the member ECs. # ===== narrowMatch: poly-specific families -- GO term applies to a SUBFAMILY only ===== - subject_id: CAZy:GT7 subject_label: "Glycosyltransferase Family 7 (beta-1,4-galactosyl-/N-acetylgalactosaminyltransferases)" predicate_id: skos:narrowMatch predicate_label: narrow match object_id: GO:0003831 object_label: beta-N-acetylglucosaminylglycopeptide beta-1,4-galactosyltransferase activity mapping_justification: semapv:ManualMappingCuration comment: >- narrowMatch: GO:0003831 describes the B4GALT-type members (B4GALT1, EC 2.4.1.38; backed by the B4GALT1 P15291 review) but GT7 also contains beta-1,4-N-acetylgalactosaminyltransferases, so the term does NOT hold for the whole family. Family alone is too coarse -> resolve to subfamily/EC before propagating (the dbCAN-sub subfamily->EC table is the route). - subject_id: CAZy:GT31 subject_label: "Glycosyltransferase Family 31 (beta-1,3-galactosyl-/GalNAc-/GlcNAc-transferases, Fringe)" predicate_id: skos:narrowMatch predicate_label: narrow match object_id: GO:0008376 object_label: acetylgalactosaminyltransferase activity mapping_justification: semapv:ManualMappingCuration comment: >- narrowMatch: GT31 is highly heterogeneous (B3GALT galactosyltransferases, B3GALNT N-acetylgalactosaminyltransferases, and Fringe beta-1,3-GlcNAc-transferases). GO:0008376 fits the B3GALNT subfamily (backed by the B3GALNT2 Q8NCR0 review, which itself proposed a more specific protein-O-mannose beta-1,3-GalNAc-T term) but not the galactosyl-/Fringe members -- the strongest case in this seed for subfamily resolution.