search

pombase / canto

The PomBase community curation tool
https://curation.pombase.org
Other
19 stars 7 forks source link

error message in existing session #1070

Closed ValWood closed 8 years ago

ValWood commented 8 years ago

worry

ValWood commented 8 years ago

couln't read genetics interactions from server error for this session. Not seen this before. Val

airvine commented 8 years ago

This sort of error usually occurs because of changes made to either the application or an ontology, such that the previously entered data is no longer compatible with the new version. For example, if an ontology term was used that no longer exists in the new version of the ontology file.

kimrutherford commented 8 years ago

It looks like an interaction in the database is missing its interacting gene. It's an easy fix, but the server is behaving oddly. Canto is accessing its data fine but I can't query from the command line without getting an error. I've copied the Canto data directory to me laptop and it's all OK here. I'll need to do some Googling to solve it.

kimrutherford commented 8 years ago

Canto is accessing its data fine but I can't query from the command line without getting an error

red face - amateur mistake: the data files are owned by the canto user on the server so I couldn't access them logged in as me. So that's a false alarm.

I've fixed the error from Canto you reported. I've deleted a mangled annotation. The annotation was for cut2 with evidence code Synthetic Lethality.

Val, could you check that no annotation is missing from the session?

I'm hoping this was caused by an old bug that has been fixed. There are two other sessions that had this problem. I've fixed them but they might need checking. Both had a mangled annotation with a missing interacting gene.

http://curation.pombase.org/pombe/curs/a434def9c7a2a3d0 - Anne Paoletti, Dosage Growth Defect, SPCC4B3.15/mid1

This one we missing the evidence too so I think the bug comes from the pre-genotypes workflow: http://curation.pombase.org/pombe/curs/a434def9c7a2a3d0 - Amanda Bird, SPAC25B8.19c/loz1

Val, could you close this when your happy with the sessions?

ValWood commented 8 years ago

In this one http://curation.pombase.org/pombe/curs/a434def9c7a2a3d0 - Anne Paoletti, Dosage Growth Defect, SPCC4B3.15/mid1 I don't see any annotation like this in the paper or in BioGRID? MAybe it was deleted?

and for this one This one we missing the evidence too so I think the bug comes from the pre-genotypes workflow: http://curation.pombase.org/pombe/curs/a434def9c7a2a3d0 - Amanda Bird, SPAC25B8.19c/loz1 it's the same session ID?

Something else is odd though:

the physical interaction in this session http://curation.pombase.org/pombe/curs/a434def9c7a2a3d0 Interactor A Evidence code Interactor B mid1 Affinity Capture-Western (affinity captures) mid1
mid1 Affinity Capture-Luminescence (affinity captures) plo1 mid1 PCA (interacts with) plo1 mid1 PCA (interacts with) rng2

don't match the annotations we see for mid1 from this paper: http://www.pombase.org/spombe/result/SPCC4B3.15#interactionPhysical affinity captures plo1 Polo kinase Plo1 Affinity Capture-Western Almonacid M et al. (2011) forms complex with rng2 IQGAP Reconstituted Complex Almonacid M et al. (2011) modified by plo1 Polo kinase Plo1 Biochemical Activity Almonacid M et al. (2011)

I am wondering if these are papers that BIoGrid and we (or in this case Anne) curated at the same time, but we both did it differently, and the confusion is to do with our biogrid synching not being quite clever enough to cope with this. I don't know if we have even thought about this situation. I know we have dealt with corrections but I can't remember whose annotations 'trump' whose. I suggest this because of the difference and also because I have never seen the evidence code PCA before, has anyone else?

We might need to discuss this ticket on Friday?

mah11 commented 8 years ago

The Amanda Bird session is bf037075feb20f28, for PMID:24831008 (I know this because it also happens to have a term request, so I looked at it for curation ticket 799). BioGRID has curated one interaction: bait adh1, prey loz1; evidence synthetic rescue. I can't tell from the session whether Amanda tried to add the same interaction or any others; it just shows the error message. I also haven't looked at the paper yet.

kimrutherford commented 8 years ago

I don't see any annotation like this in the paper or in BioGRID?

It could be that she started making an annotation as a test and didn't delete it. I'm now sure it was a half finished annotation.

mah11 commented 8 years ago

I'm going to email Amanda soonish anyway, once I've added the FYPO term for that session, so I can ask about the interaction annotation too.

I won't mess with anything in the Anne Paoletti session ;)

ValWood commented 8 years ago

I'll deal with Anne's session, will keep this ticket open....I'm still a little confused here.

ValWood commented 8 years ago

PMID:2203537 (cut1 session) fixed

ValWood commented 8 years ago

I think possibly the A Paoletti session has not yet been submitted to curators? I think that is why it has the existing biogrid annotations and Anne's new ones are slightly different. If this can be confirmed, there is need to check this as it will be checked when it is completed and submitted. val

kimrutherford commented 8 years ago

I'll deal with Anne's session, will keep this ticket open....I'm still a little confused here.

Hi Val. Have you been able to sort this out. Can I help?

ValWood commented 8 years ago

Re Anne's session, I'm not sure where it is up to so I mailed her to see if she had submitted it. Once we get it through to approval we can check the details. I'm not sure if we need this ticket or not. I suspect it can be closed?

kimrutherford commented 8 years ago

Yep, let's close this. The original problem is fixed.