Manager Call 2019-08-07
- Agenda: Huaiyu
- Minutes: Suzi
GO users meeting Berkeley 2019
Mail from Nomi:
We need to firm up plans for the GO User Meeting (Oct 7). I’m kind of out of the loop on this (though I’m working on logistics). Who is the intended audience, and how are we advertising to them? How many people do we expect to attend (and how many of those will be external users rather than GO project members)? Seth said that the last GO User Meeting had very few external attendees. Seth said that AGR people might be visiting on Oct 7. Is it possible that Oct 7 therefore might end up being a GO/AGR collaboration meeting rather than a User Meeting? If this topic needs wider discussion, feel free to forward this message and/or let me know if you’d like to devote some time to it during the next GO Managers (?) call. Thanks, Nomi
GO release stats
Huaiyu, Chris, David, Seth, Suzi, Pascale, Paul, Kimberly
Planning: Montreal was last minute, with few last-minute and a few extra people coming
Hoping to get KBase and enigma (?) to come, JGI too. These presentations will be for a broader audience.
What is space rat: exposing rats to radiation similar to what expected in space.
Bringing anyone else in- no plans yet for tech or pharma, Berkeley who to target- algorithm/tool developers, general users, students/post-docs,
Do we have training tools PT- yes for EA Pascale: some from montreal and DC
Seth- Need a deadline to plan efficiently enough for this to happen
Alliance developers invited? Yes, not as useful for them to be there during general users meeting
Pascale- announce to users where to use files, how to use GO
Logistics- if this is LBL, how to target students/postdocs. Can make dev-focused, but then need to tarteg devs to attend Need to advertise either way- recruit people to come. Difficultly in convincing people to travel, local easier. Need to commit.
User meeting may be more appropriate as an ISMB piggyback Some devs- Dexter (Chris M) v eager to come up. This would be more targeted than an ISMB session.
Is there enough of a critical mass to do this kind of meeting with locals (5 too small, 20 would be worthwhile) Could use this as an opportunity to develop materials for this kind of thing
Informal would be fine, would like to spark developers- to use GO-CAMs etc. This would be a really good audience, alternate to traditional "end users meeting"
CM- needs a few says to poll users
SA- to gather materials used in Montreal, etc in case we do an end-users meeting at 'the drop of a hat', to be deposited _________ (no existing place, GO End Users Meeting Drive?) Everyone to drop what they have
Agenda: Revisit, make final decision on what type of meeting to have.
Project Status (Kimberly)
Most recent effot is toward ShEx specs and accompanying google docs specs. Specify what is allowed in GO cam model, basis for validation checks. Also specify what is allowed in MOD imports, extensions, development of Noctua Form and annotation review.
Call later today (3ET) to look over status of both, particularly what is there for MF and contextual info about MF.
Noctua Form and annotation review form- resumed Tuesday calls (seth, tremaine, LP, KVA, Pascale). Form 2.0 and tickets being addressed specifically.
ShEx specs are priority over import; hope to finish around October. Will ensure imports are compatible.
Need discussion with Alex about which extensions are allowed in Protein2GO.
GO Release Stats (seth)
LP put script that works with Solr, much more detailed stats (changes in ont, in annotation numbers) not yet running
No stats as part of release, but there are stats from a pipeline. Hoping to restart releases in next few days- beginning of next week and stats will be checked by a few people.
Idea: distribution of annotations to slims
Pathways 2 GO (David)
Goal: import data from Reactome into GO Cam models with great fidelity (ben lead, others give instructions as needed)
Project in GO GH as Pathways2GO
Pathways exported into BioPax, then applying rules to convert to GOCAM model. Successful with a few minor burps.
Example: BMP pathway has been teased apart. Most issues have to do with GO vs Reactome views of world/how they do things.
Is a manuscript in Google Docs including what is lost and compromises made.
GABA, glycolosis, a few other pathways extensively worked on and made into line models; works as QC for Reactome and GO CAMs
Reactome describes sets of molecules or complexes as having functions; Ben made ontology of Reactome and these are used in GO CAMs- can now be converted into classes and individual molecules assigned to pathway
Lines became complicated, went to cytoscape view. Reactome is reaction-based, not as important in GO so outliers in pathways important in Reactome are represented but certain relations will (in future) outweigh secondary relation shown in GO CAM (not be part_of primary process in GOCAM)
That leads to ability to connect go cam models in future.
Reactome has been re-curated after some pathways mapped in GO CAMs
PT: Need standard ways to read go cams- left to right, top to bottom preferred. Also the MF would be useful information to add to Reactome info- is MF not known or just not represented (answer: not represented in DH's slides)
Good opportunity for co-curation between groups