Manager Call 2019-11-6
- Agenda: Laurent-Philippe
- Minutes: Kimberly (in italics)
Projects Specification guidelines
- Start to spec out projects. Right now there are too many high-priority things. We modified the template document on the Google drive.
- There is no longer a need to make a GH project for a project that is not yet active.
First pass at project priorities and PIs will assign the Product Owners / Tech Leads, especially green tasks (there are 42 high tasks). Meanwhile, reviewing all high / green tasks and assigning very high categories for anything related to the Noctua 2.0 / GO-CAM main delivery.
Main objective for the priorities spreadsheet is to make sure that we have product owners and tech leads for each project of very high priority. The product owners represent the users and the tech leads determine how software will be implemented to meet the stated objectives. If there are concerns about the people assigned to each project, please bring this to the next managers call.
- Potential blocking issue of wanting a new relation to replace existing relations that are currently in the ontology for terms like 'cell fate specification'.
- This, along with other spatial relations, e.g. 'occurs adjacent to', will require some work on RO.
- We'll need to devote some resources to working on RO to accomplish these tasks.
- Annotation history also set to high priority as this is required for MOD imports, hence for the main delivery.
We will import MOD annotations with as literal translation of the existing annotations as possible. This means we'll be keeping the 'results in' relations in the imports for now. We do want to address these relations in the future, but don't want this work to block the imports. The MGI and WB annotations are nearly all ready for import.
The SAB report was emailed to go-managers.
We'll discuss the SAB more fully next week when more people are on the call.
In some cases, AmiGO-loaded ontologies seem sub-optimal, as it returns no-GO terms (often more than GO terms) https://github.com/geneontology/go-ontology/issues/18048
- Non-GO terms should NOT be returned in the search results
- Obsolete terms should be weighted down (or an option to not get obsolete terms in the search results)
- What is the priority on this task ? Considering people write that they are using QuickGO because of this (and other usability issues)
The behavior of AmiGO with respect to non-GO terms needs to be spec'ed out for future work. It's unlikely that there's a quick UI fix for the search or tree view issues raised by Leonore and others.
PRO gpi file
- Harold reports from the PRO meeting (week of October 21st) that PRO is now providing a production gpi file released every 3-4 months.
- PRO is also providing a 'development' file with more up-to-date additions to accommodate the needs of MODs that use PRO for their curation and want to include PRO ids in their respective gpi files.
- We need to establish an SOP for id/accession dataflow between PRO and GO annotation groups.
- We don't want duplicate ids in gpi files from PRO, MODs, and UniProt.
- How do we make annotations that curators make with PRO ids available?
- Note that there is a more general issue wrt allowed entity identifiers as primary annotation objects, With/From entries, and annotation extensions.
- We need to deal with this issue, too, since it is creating problems downstream for consuming groups and may have implications for what we put into NEO.
We need to develop clear specifications and guidelines on what ID spaces are allowed for a given organism, how NEO should handle those ID spaces, and what should be available for curators to use for annotation. Once the ShEx specifications are finished (at least this first round), this might be a good topic for future GO-CAM specifications calls.
GO Out-of-Office Calendar
Pascale: I’ve added people in the managers team in my ‘GO people away’ Google calendar. Let me know if you haven’t received the invite or if you want another email address to be associated.
Contact Pascale if you do not have edit access.