red tile fish (ncbi taxon:270562)

[selewis]

Annotation to GO:0006265 (DNA topological change) should be true

Tested with http://owlservices.berkeleybop.org/isClassApplicableForTaxon?format=txt&idstyle=obo&id=GO:0006265&taxid=NCBITaxon:270562

[selewis]

This looks like an error upstream of us in the PANTHER or UniProt data where BRAJA was used instead of the correct 5-letter code BRAJP

The node claiming to be of taxon 270562 would make it BRAJA, aka Branchiostegus japonicus (the red tile fish) In fact it looks like this node should be taxon 375 that is, BRAJP Bradyrhizobium japonicum This is correct both in terms of the branch it is in and the species label (which is Bradyrhizobium japonicum)

[selewis]

Message sent Digging a bit deeper this looks like a mistake upstream. Either in PANTHER or UniProt.

The node with the mistake is AN618:BRAJA|Gene=BAC46088|UniProtKB=Q89W70 in family PTHR10169 (aka PTN000018877)

The 5 letter code given to this node is BRAJA, when it should be BRAJP, hence the taxon lookup was coming back with the red-tile-fish rather than the proteobacteria it should be.

Huaiyu and/or Paul - can you find where this error is cropping up?

How quickly can you change the data that provided when we load this family? A horrible hack would fix it temporarily, but this could cause other people problems too.

[selewis]

Hack in, but now the taxon server doesn't apparently know about ncbitaxon:375 for BRAJP

[cmungall]

Currently the taxon server loads ncbitaxon/subsets/taxslim, which is guaranteed to contain the uniprot reference proteome. There is no taxon 375 in ftp://ftp.uniprot.org/pub/databases/uniprot/current_release/knowledgebase/reference_proteomes/README

[selewis]

Yes, but on that page they do have both UP000002526 224911 BRADU 8024 229 8314 Bradyrhizobium diazoefficiens (strain JCM 10833 / IAM 13628 / NBRC 14792 / USDA 110) UP000001994 114615 BRASO 6551 150 6717 Bradyrhizobium sp. (strain ORS278) Which are related. PANTHER guys, have you done some mapping here?

[huaiyumi]

I took a look. This is an issue dating back to 2010 when PANTHER v.7.0 was released. The code BRAJA was used in that release, which was for Bradyrhizobium japonica. It should be BRAJP. BRAJA is for a different species (a fish). Another problem compounded to this is that the taxon ID we used for this code was 224911, which is for Bradyrhizobium diazoefficiens. Therefore, we have been using Bradyrhizobium diazoefficiens genome in our trees but called it Bradyrhizobium japonica, and used the incorrect code. This will be corrected in our next release. We won't be able to fix it in v.10.