pgaudet
commented
1 year ago Some strange examples:
1) RGD 61865 Lbp GO:0033036 PMID:15294986|RGD:9068941 ISO UniProtKB:P18428 P lipopolysaccharide binding protein gene NCBITaxon:10116 20170221 AgBase CHEBI:46889|CHEBI:134411|UniProtKB:P08571
- this is not in AmiGO -current although it's marked as have been annotated in 2017
- the annotation was deleted on 19-Sep-2022, so it's odd that it would appear in this next release
2) RGD:1303008 mutS homolog 5 GO:0000795 "{""relationship"": {""relation"": [{""id"": ""BFO:0000050"", ""label"": ""part of""}], ""id"": ""CL:0000017"", ""label"": ""spermatocyte""}}|{""relationship"": {""relation"": [{""id"": ""BFO:0000050"", ""label"": ""part of""}], ""id"": ""EMAPA:17972"", ""label"": ""EMAPA:17972""}}" RGD NCBITaxon:10116 ISO MGI:1329021 PANTHER:PTHR11361 gene RGD:1624291 20060817
This annotation is not in MGI: https://www.informatics.jax.org/go/marker/MGI:1913739
In AmiGO staging, while RGD lost 17k ISOs, it now has a lot more sources of ISOs:
Current: 225895 RGD 133 UniProt 117 ComplexPortal 34 ParkinsonsUK-UCL 33 BHF-UCL
Staging: 207344 RGD 1127 UniProt 200 ComplexPortal 151 HPA 115 BHF-UCL 81 CAFA 75 MGI 73 ARUK-UCL 56 ParkinsonsUK-UCL 21 CACAO 14 AgBase 11 SynGO 5 HGNC-UCL 2 FlyBase 1 DIBU 1 GO_Central 1 IntAct 1 Reactome 1 WB
However I doubt very much that the ISOs were made by the sources cited, since many of these groups do not produce annotations anymore. Also, the original reference is kept in most cases (but that was already the case, AFAIK).