The QuickGO and UniProt websites are not responding at the moment, so I can't see where the Bacillus subtilis annotation is coming from, but it sounds like an issue with (a missing) taxon check during QC procedures?
Thanks,
Paola
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Sorry, clearer now that the QuickGO page loaded fully: the Bacillus annotation is indeed a PAINT one, so PAINT might want to look into their pipeline/implementation of taxon checks in general. Thanks!
Paola
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Yes, Paola is correct and I should have been clearer in the report. It's a problem with PAINT not catching the taxon constraints that the terms already have, which is why I reported it here instead of on the ontology requests tracker. I found this when a user asked for help extracting a B. subtilis GAF... it just happened to be the first thing I saw in AmiGO when I searched genes and gene products for Bacillus.
I vaguely remember seeing a similar problem with another PAINT annotation that we imported into the E. coli GAF. In that case the GAF submission QC caught it, so I guess there is also a problem with PAINT annotations not being tested there... or is the file with all of UniProt not checked because it's too big?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Jim,
This is a family that was curated a long time ago, before the taxon constraints were implemented. I reopened the family on my machine and will commit a cleaned version.
We do have reports of a lot of those errors. We should discuss with Chris and other if these annotations that violate the taxon constraints should be filtered put of the GO database. Right now we get an error report, and we need to manually correct them.
Pascale
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Yes, we should filter out violations as part of the GAF publishing
pipeline (pre loading into GO database)
Can be discussed next week at these junctures:
Tuesday
* Phylogenetic Annotation: Jenkin running of PAINT GAF file updates
(Suzi)
* Annotation Workflow: Jenkins GAF QC checks and inferences (Heiko,
Rama)
Wednesday
* Software infrastructure - part 2: Migration of all cron jobs to
Jenkins - status
On 9 Oct 2014, at 10:38, Pascale Gaudet wrote:
Hi Jim,
This is a family that was curated a long time ago, before the taxon
constraints were implemented. I reopened the family on my machine and
will commit a cleaned version.
We do have reports of a lot of those errors. We should discuss with
Chris and other if these annotations that violate the taxon
constraints should be filtered put of the GO database. Right now we
get an error report, and we need to manually correct them.
Hi Jim,
GO:0005739
mitochondrion
already has a taxon rule: only in Eukaryotes
relationship: only_in_taxon NCBITaxon:2759 {id="GOTAX:0000060"} ! Eukaryota
The QuickGO and UniProt websites are not responding at the moment, so I can't see where the Bacillus subtilis annotation is coming from, but it sounds like an issue with (a missing) taxon check during QC procedures?
Thanks,
Paola
Sorry, clearer now that the QuickGO page loaded fully: the Bacillus annotation is indeed a PAINT one, so PAINT might want to look into their pipeline/implementation of taxon checks in general. Thanks!
Paola
Yes, Paola is correct and I should have been clearer in the report. It's a problem with PAINT not catching the taxon constraints that the terms already have, which is why I reported it here instead of on the ontology requests tracker. I found this when a user asked for help extracting a B. subtilis GAF... it just happened to be the first thing I saw in AmiGO when I searched genes and gene products for Bacillus.
I vaguely remember seeing a similar problem with another PAINT annotation that we imported into the E. coli GAF. In that case the GAF submission QC caught it, so I guess there is also a problem with PAINT annotations not being tested there... or is the file with all of UniProt not checked because it's too big?
Hi Jim,
This is a family that was curated a long time ago, before the taxon constraints were implemented. I reopened the family on my machine and will commit a cleaned version.
We do have reports of a lot of those errors. We should discuss with Chris and other if these annotations that violate the taxon constraints should be filtered put of the GO database. Right now we get an error report, and we need to manually correct them.
Pascale
Yes, we should filter out violations as part of the GAF publishing
pipeline (pre loading into GO database)
Can be discussed next week at these junctures:
Tuesday
* Phylogenetic Annotation: Jenkin running of PAINT GAF file updates
(Suzi)
* Annotation Workflow: Jenkins GAF QC checks and inferences (Heiko,
Rama)
Wednesday
* Software infrastructure - part 2: Migration of all cron jobs to
Jenkins - status
On 9 Oct 2014, at 10:38, Pascale Gaudet wrote:
Related
PAINT:
#66I have fixed the offending family. I know we have other taxon constraints violations, we need to check them all.
Can this issue be closed ?