Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Apr 16 16:37
    pgaudet unassigned #3736
  • Apr 16 16:37
    pgaudet edited #3736
  • Apr 16 16:36
    pgaudet edited #3736
  • Apr 16 16:36
    pgaudet unassigned #3736
  • Apr 16 16:36
    pgaudet commented #3736
  • Apr 16 16:36
    pgaudet commented #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:33
    pgaudet assigned #3736
  • Apr 16 16:29
    pgaudet opened #3736
  • Apr 16 16:25
    pgaudet edited #3735
  • Apr 16 16:24
    pgaudet edited #3735
  • Apr 16 16:10
    pgaudet edited #3735
  • Apr 16 16:10
    pgaudet edited #3735
kltm
@kltm
@balhoff Was that for production or dev?
Jim Balhoff
@balhoff
production
sabrinatoro
@sabrinatoro
thank you!
Eric Douglass
@dougli1sqrd
Hi @kltm, @pgaudet, @cmungall, So in ontobio, the tests for qualifers seem to indicate that involved_in should be allowed. But http://geneontology.org/docs/go-annotations/#annotation-qualifiers says only NOT, contributes_to, and colocalizes_with are allowed. I'm making corrections in ontobio to catch the space issue, but I'm not sure how to proceed given the new information here
kltm
@kltm
@dougli1sqrd
It only appears twice:
goa_uniprot_all-src.gaf.gz:2
mgi-src.gaf.gz:12
It may be worth talking to @ukemi
kltm
@kltm
Especially if we are to start booting them, @ukemi should be in the loop.
pgaudet
@pgaudet

So in ontobio, the tests for qualifers seem to indicate that involved_in should be allowed

I am not clear what that means. Surely ontobio is not where the specs are coming from ?

We have decided to add other relations in the GAF (including involved_in, and others), but this is not yet the case, so for now all qualifiers other than NOT, contributes_to, and colocalizes_with should be filtered out
sabrinatoro
@sabrinatoro
In the delta-notch pathway: is delta part of the pathway, or is it regulating the pathway? Is there already an agreement on this?
Thanks
Helen Attrill
@hattrill
Hi @sabrinatoro, as the definition includes the ligand, it should be part of the pathway.
sabrinatoro
@sabrinatoro
@hattrill thank you for confirming!
sabrinatoro
@sabrinatoro
Quick questions:
1- what is the update about the users meeting on Monday? Is it happening?
2- It said that we should receive an email with the link to join the GO-meeting. Has this email already been sent?
Thanks!
I'm not sure what is going on with an email at this point.
sabrinatoro
@sabrinatoro
Thank you Seth!
kltm
@kltm
My guess is that it was only distributed to people who had "signed-up".
sabrinatoro
@sabrinatoro
I did sign up and I haven't received anything... Just wandering whether it is 'normal' (i.e. hasn't been sent yet), or I 'fell through the cracks'.
kltm
@kltm
Ah, it looks like it may be getting sent out later today. Let me check on that.
It looks like you should be getting it later today then.
If you do not, ping us and we'll make sure you get the channel/password.
sabrinatoro
@sabrinatoro
OK. Thank you for looking into it!
Eric Douglass
@dougli1sqrd

@dustine32 @vanaukenk I'm looking at the gpad 2.0 spec. For with/from: Pipe-separated entries represent independent evidence; comma-separated entries represent grouped evidence, e.g. two of three genes in a triply mutant organism

We currently only model with/from as a list. Could/should the pipes be treated as different annotations? Like if I had withfrom a|b, this would be the same as 1:a, 2:b? Do we need to update how this is modeled internally?

vanaukenk
@vanaukenk
@dougli1sqrd @dustine32 @ukemi I believe that for the imports Dustin has modeled pipe-separated With/From entries as separate annotations.
kltm
@kltm
It will be good to get on the same page. @vanaukenk If we then output these separate annotations as separate lines in GPAD output, would that weird people out?
dustine32
@dustine32
@dougli1sqrd Yep, gocamgen should be splitting pipe-separated with/from values into separate GO-CAM assertions
vanaukenk
@vanaukenk
@kltm No, I don't think it would weird people out too much. The piped with/from field was really more of a convenience and a way to reduce the number of annotations lines, rather than anything to do with the actual data, iirc.
kltm
@kltm
Okay, that's good to know. It's good to know that intention is maintained.
Eric Douglass
@dougli1sqrd
cool cool thanks everyone
Eric Douglass
@dougli1sqrd
What is the difference between a qualifier and a relation? It seems like we kind of mix and match these
kltm
@kltm
Well, it depends. Have a minute?
Loosely, a qualifier is a thing that "qualifies" the gp2term relation-ish thing; a thing found in that column in GAFs. It could be NOT or a "relation".
Eric Douglass
@dougli1sqrd
Prob not, can we chat after software?
kltm
@kltm
NP.
A relation is something you might find in RO.
The problem with how we used these words is that they evolved through time, starting from when the GO was a lot less strict.
The definition of GPAD in BLML and GO-CAM hopefully does away with any remaining ambiguity.
pgaudet
@pgaudet
Humm. I thought the qualifiers were NOT, contributes_to, and colocalizes_with, and all others were relations. @vanaukenk ?
kltm
@kltm
@pgaudet Technically, that's RO:0002326 and RO:0002325, but if we're thinking about history it's a little fuzzier as we move back in time. The question is originating in thinking about the data model that's being used for parsing (I think we've sorted it out in another channel).
suzialeksander
@suzialeksander
Speaking of ROs, I’m looking at the new GPAD specs and have a question about extension formatting. In the GAF, we had part_of(GO:0071474),has_input(UniProtKB:P32485) In GPAD, would that be BFO:0000050(GO:0071474)|RO:0002233(UniProtKB:P32485)? Also, I think has_input is OK even though it’s not on the specs? @vanaukenk @kltm
kltm
@kltm
@suzialeksander Maybe a @vanaukenk question if there needs to be clarification, but if it's not in the spec it's verboten (and maybe there needs to be some discussion). The gp2term list is explicit (col 3), but the extension field allow everything from RO and go_rel.
Your extension field looks right to my eye.
From the spec, for the extensions:
The Relation_ID may be from the Relations Ontology or from the set of GO relations, go_rel.
suzialeksander
@suzialeksander
OK, I think I’m confusing the “what’s allowed in col 3” with “what’s allowed in the gpad”. That makes sense
kltm
@kltm
Yeah, maybe that could be made more clear: there is col 3 and col 11.
Col 3 is the table and col 11 is a free-for-all*.
suzialeksander
@suzialeksander
Yep, I just mined the RO from oborel /obo-relations on Github
excellent, on 11 being a free for all
kltm
@kltm
I mean, there's that "*", but yeah, I think it works like you're thinking.
suzialeksander
@suzialeksander
I have no idea what the asterisk is for, I’m assuming it’s some terms&conditions I’m tacitly agreeing to by reading the specs...