Closed ValWood closed 6 years ago
another "Gotcha"
I've removed dubious and transposon genes and excluded NOT annotations.
Those changes bring the count down to 1005.
I'm still working on adding the counts and links for the non-slim genes to the go slim page.
I've removed dubious and transposon genes and excluded NOT annotations.
Did this work only partially?
Note that I am now adding an explicit annotation to "transmembrane transport". Its a transmembrane transporter, just not biotin, so it WILL slim to TM transport tomorrow. v
OK, this (SPAC1B3.15c) already has
/GO="aspect=P; term=transmembrane transport; GOid=GO:0055085; evidence=ISM; db_xref=GO_REF:0000050; with=Pfam:PF07690; date=20160526"
so the problem seems to be that the term is recognsed, and SPAC1B3.15c is slimmed appropriately. However, "transmembarne transport is not displayed (presumably the NOT is ignored for filtering)
OK, this (SPAC1B3.15c) already has
Looks like there are two bugs. :-(
The transmembrane transport annotation is missing from Chado. Maybe there is a problem with the GO filtering and NOT annotation.
But a slim term is showing anyway so look like there's a web problem too.
I'll work on that.
Maybe there is a problem with the GO filtering and NOT annotation.
Turns out that we have an very old issue about this that we punted to the future: https://github.com/pombase/pombase-chado/issues/35
I guess we need to think about it?
But a slim term is showing anyway so look like there's a web problem too.
I've fixed that problem at least. The web code now correctly handles NOT annotations. We still need to fix the GO filtering.
http://preview.pombase.org/gene/SPAC1B3.15c needs to map to "transmembrane transport " in the slim
If anything else needs to happen here please update the ticket. If it isn't clear add "discuss". It isn't critical for release, and I don't know of another example. I will add the quick label here without the milestone.
Turns out that we have an very old issue about this that we punted to the future: pombase/pombase-chado#35 closing as duplicate
That will probably do it. The other issues are related to GO, or curation. I'm working on reducing the number which do not map to the slim today. Should improve tomorrow.