Participants: Midori Harris, Jennifer Clark, Rex Chisholm, David Hill, Eurie Hong, Judy Blake, Jane Lomax, Chris Mungall.
Agenda/Chair: Chris Mungall
Minutes: Eurie Hong
- 1 Review Action Items from previous meeting
- 2 Individual Manager reports
- 3 Meetings
- 4 Collaborations and Interactions, active and new
- 5 Operations and Procedures
- 6 Staffing and Personnel
- 7 Budget issues
- 8 Publications
- 9 Other Current Topics and Concerns
- 10 Summary of Action items
Review Action Items from previous meeting
Chris: Give webex demo of tool for mining biochemical pathways from reactome?
The original intent of the action item was for Chris to be involved in a WebEx meeting with the biochemists who will be involved in doing the links between function and process. Chris did take use the Reactome BioPax format and the reactome2go mapping in order to do an intial assessment of the mapping and to determine if the part_of relationships are maintained between GO and Reactome. The mapping is not a direct 1:1 mapping. This type of mapping between Reactome and GO will also need to be done for other types of related data. There will be a lot of manual review of the mapping which will serve as an impetus for improving both groups data. Action Item: Chris to set up WebEx with folks working on this project to set the stage and get a scope of the project.
Evelyn Camon, Judy Blake, Chris Mungall, Jennifer Clark to explore possible cancer grant proposal.
Jen will send out the URL again.
Anyone who wants to help the ontology development working group:
- What are your areas of biological expertise? We will be calling on everyone to join the small conference calls/content meetings and it would be very useful to have a clear idea of which topics each individual can contribute to.
- When do you have time to contribute? Each person has quiet periods and more busy times when they may be traveling or writing grants. It would be very useful to know when people are less busy so we can call on them to contribute. Please mail Midori and David.
All: Has any member of GOC been consulted on the Biosapiens project?
Michael Asburner is working on this.
PI's: Consider if we should send someone to MGED.
Jen: ask Fiona for her thoughts on this.
Still an action item.
All: Update your godatabase links.
Individual Manager reports
- The group is still working on a flowchart for documentation.
- Jen often gets asked to talk at meetings but then needs to talk about annotations. Often it is not clear that this is what the organizers want. Jen is being trained by Emily to do annotation and will be working on annotations in the morning for the month of June.
- Plant sensu: continuing to work on this
- Transporter function ontology: continued progress
- MIT meeting : David is going to write up the summary of the meeting. GO PIs are also having a conversation about future directions. Michael @ MIT is working on computing on the multi-cellular process, cellular process, and interaction between organisms branches. The upshot is that the results need to be reviewed biologically because sometimes it makes a lot of sense why those terms called "too general" need to be there. Sometimes it's a by-product of annotation.
- Regulation terms: A crux of the issue it that there needs to be a definition of the 'regulates' relationship.
- ChEBI: We have a lot of true path violations for things like 'serotonin' which acts as a neurotransmitter and a hormone in the immune system: a term 'serotonin transport' is_a child of 'neurotransmitter transport'. However, we cannot use ChEBI because their "is_a" relationships different from the GO "is_a" relationship because ChEBI allos is_a to mean "sometimes can play a role".
Action Item: Jane to set up a conference call with the folks at ChEBI to begin working with them so ChEBI can align better with GO.
- figs2go mapping: Folker Meyer from the SEED project came to visit about mapping to GO. There is overlap with the IMG (JGI) mapping. Jane has made this her top priority.
- ontology versioning: Chris brought this up in response to an inability to reproduce results that are published. Users need to be able to define which version of the ontology they downloaded and used. The CVS versionining is specific to the file itself. This in and of itself can lead to confusion. The versioning for the gene_ontology_edit.obo file may be different than the gene_ontology.obo file which may be different than the version that is checked out of the anonymous CVS.
We would also want the databases to be loaded with the same versions that are available for downloading.
Action Item: Chris to write up a summary of the issues and reasons to present at an upcoming manager's call.
- Interaction with Reference Genome group: Mary Dolan and Sohel Merchant will work with the reference genome group to deal with data storage issues. The Google spreadsheet is a nightmare. And work with the AmiGO working group to deal with display needs. There are several issues that need to worked out in order to do this:
- The incorporation of orthology groups. Rex has been in touch with Richard Durbin to incorporate TreeFam as a companion to the information provided by YOGY. One issue is that TreeFam uses transcript IDs which can have many to 1 mapping to a gene at a MOD.
- ID mapping. This is actively being worked on by the software group.
- Range of release schedules in the group. There are groups who work on the live database and groups who release data in month-long or more cycles.
- OBO-Edit version 1.1 is nearing release. Need signing off from OBO-edit working group.
- Adressed above
- Testing an AmiGO with IEAs: User experience is a bit sluggish but still need actual analysis of how slow the database is with IEAs. And then explore views of the data.
Judy will be attending ISMB.
Collaborations and Interactions, active and new
Operations and Procedures
Staffing and Personnel
Jane was wondering whether there should be a publication addressing how to develop and modify ontologies. More of a process paper. It could be split into multiple domains - the technology aspect and the biology aspect. Like focusing on a content meeting for the biology. There is a publication by Jonathan Bard and Sue Rhee in Nature Reviews Genetics (PMID: 14970823) that discusses ontology design.
There was some discussion on the CNS content meeting bringing written up that included the analysis that Erika Feltin is wanting to do on the ontology.
Another suggestion was putting the process as an FAQ type question with a list of relevant publications.
In addition, there was a suggestion for writing up the process as an introspection of how GO has changed. Possibly as part of a 10 year anniversary paper for GO.
Other Current Topics and Concerns
Summary of Action items
- Chris and the biochemists involved in making links between function and process to set up WebEx with folks working on this project to set the stage and get a scope of the project.
- PIs: should we send someone to MGED?
- Jane to set up a conference call with the folks at ChEBI to begin working with them so ChEBI can align better with GO.
- Chris to write up a summary of the issues and reasons in order to incorporate an ontology versioning system to present at an upcoming manager's call.