Open ukemi opened 2 years ago
Related to #424
From 2021-10-05 MOD Imports call:
It looks like the 'located in' annotations are coming from annotations that originally use 'part of' but for which the gp2term relation is being corrected by GO_RULE:0000061 to 'located in'.
The 'part of' annotations are generated by a property chain 'enable' o 'occurs in' -> 'part of' that is coming from one of the ontologies imported into GOLEGO.
@balhoff will look at GOLEGO in Protege to see where the chain is coming from and then update the GOLEGO Makefile so we can filter things like this.
I believe I fixed the issue in go-lego. The bad SWRL rule should be eliminated in the next snapshot build.
Thanks @balhoff So was it the case that the errant SWRL rule was in an old version of RO and we needed to update the RO version that GO uses in the ontology to fix go-lego? Just trying to understand the ontology data flow here. Thx.
There were two issues:
Okay, great; thank you! Should WB then review and update its WBLS file?
@vanaukenk it won't impact us now, but I assume they would want to update their RO import for the next release. Maybe that will happen normally anyway?
Do the SWRL rules exist in a specific location? It would be good to add this a something to check when we change anything about relations.
They are part of RO, but sometimes also we will get duplicate axioms from ontologies that import RO, like with the WBLS issue. This is mainly addressed by using input files for those ontologies that exclude their external axioms.
@ukemi - do you know if this issue is fixed for the MGI GPAD?
It wasn't the last time I checked.
Okay, then we can close this ticket? I am double-checking about the WB ontology files, but Jim has addressed that on GO's end.
Note that the issue with the WBls file was likely a timing issue wrt when the updated RO file was generated and the WBls file used in go-lego was generated.
WBls has now been updated with the most recent RO, so should be okay.
I am still seeing this in the snapshot QC report:
GAF lines are parsed according to GAF 2.1 specifications
total: 1169
Messages
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:1098687 Aak1 is_active_in|located_in GO:0098793 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1305520 C AP2 associated kinase 1 5530400K14Rik|D6Ertd245e protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:106388 Abhd17a is_active_in|located_in GO:0098978 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1359682 C abhydrolase domain containing 17A 1700013O15Rik|D10Bwg1364e|Fam108a protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:106388 Abhd17a is_active_in|located_in GO:0099031 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1359682 C abhydrolase domain containing 17A 1700013O15Rik|D10Bwg1364e|Fam108a protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:106388 Abhd17a is_active_in|located_in GO:0099033 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1359682 C abhydrolase domain containing 17A 1700013O15Rik|D10Bwg1364e|Fam108a protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:1917816 Abhd17b is_active_in|located_in GO:0098978 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1305246 C abhydrolase domain containing 17B 5730446C15Rik|Fam108b protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:1917816 Abhd17b is_active_in|located_in GO:0099031 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1305246 C abhydrolase domain containing 17B 5730446C15Rik|Fam108b protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:1917816 Abhd17b is_active_in|located_in GO:0099033 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1305246 C abhydrolase domain containing 17B 5730446C15Rik|Fam108b protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:1917428 Abhd17c is_active_in|located_in GO:0098978 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1308210 C abhydrolase domain containing 17C 2210412D01Rik|Fam108c protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:104913 Abi1 is_active_in|located_in GO:0014069 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:621008 C abl interactor 1 E3B1|Ssh3bp1 protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:104913 Abi1 is_active_in|located_in GO:0098978 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:621008 C abl interactor 1 E3B1|Ssh3bp1 protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:107771 Abr is_active_in|located_in GO:0098978 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1306279 C active BCR-related gene 6330400K15Rik protein taxon:10090 20180711 MGI
ERROR - Invalid Qualifier: There can be only one relation entry for the Qualifier field (is_active_in|located_in) -- MGI MGI:107771 Abr is_active_in|located_in GO:0099092 MGI:MGI:4417868|GO_REF:0000096 ISO RGD:1306279 C active BCR-related gene 6330400K15Rik protein taxon:10090 20180711 MGI
Spotted as part of the MGI imports. From more complicated GO-CAM models we are generating these two annotations:
I think these would fall into our (the GOC's) definition of redundant. @loricorbani