AHWG-20070626

From GO Wiki
Jump to: navigation, search

Chair: Rama
Present: Seth, Eurie, Amelia, Jane, Chris, Ben
meeting mode: Phone call


  • About meeting Chairs- No body in the group (other than Ben and Rama) was aware of the reverse order of Chairs for Hub and WG meetings. We decided we will keep it simple and the next person in the list will chair the next meeting (we won't differentiate Hub and WG meetings). We will put a Star next to the name who is going to chair the next meeting.
Action Item_ Ben will update the Wiki about Chair rotations
  • Main agenda- We went through the 'big features' that Chris sent around last week.


On Jun 21, 2007, at 6:26 PM, Chris Mungall wrote:

Here are some non-trivial new features I'd like to see added to AmiGO over the next year. These are kind of orthogonal to features like shopping carts, which are needed to. They often involve new types of data, or radically new displays

  • bi-axial grid type exploration of annotations (eg BP/CC, or BP/Taxon). See the emails I sent a month or so ago.


We will make this an optional view, not default. For people doing advanced queries. How easy would it be to implement? Chris thinks it shouldn't be hard. This view can be used to generate a slim. No time frame yet. Will be useful to have a demo for the WG.

We will make sure we take care of items from the last list before the next Consortium meeting (references, structured comments, reference genomes). For references genomes, we need the orthologs in the DB. Then Seth/Amelia can take over. That is going to be discussed in the next software meeting.

  • option to split a tangled DAG into separate is_a and part_of views (and other relations, like regulates). As in oboedit

(requires improvements to graph_path building)


Will be discussed over email.

  • combined function-process views; pathway style?

(relies on links being established)

The biaxial grid will help to figure out these. We need some mapping to proceed on this.

  • cross-product displays

(can test with current GO-CL cross products. May require API changes)

  • advanced inter-ontology queries

(example: what kinds of processes potentially gives rise to nurse cells? this can be answered if we know that cystoblast division results_in_division_of cystoblast, and nurse cell develops_from cystoblast) (depends on xps)

Will require loading the cell ontology into the DB. Has to be discussed in the larger meeting with some mockups. We don't want to answer user questions about other ontologies. So, we won't show/allow to browse the other ontologies, but we can integrate it.

  • orthology and protein family based views

In addition to ref.genomes. chris will look at the grant proposal to see what we proposed for ortholog displays.

  • Facebook style API to allow people to develop their own applications that plug into amigo, can be shared with friends... particularly for doing specific kinds of analyses. In addition, I'd like to use AmiGO for other kinds of OBO annotation - this isn't something GO is funded to support. But it would be nice to reuse code as much as possible.

Amigo infrastructure will allow that.

Chris will put these ideas on the tracker with more detail/descriptions.

Eurie- these are very sophisticated ideas. We should prep the WG by just going through these ideas. Many of them are not aware of the changes going on with the ontology file and they don't know how everything comes together. If it is easy to do a mock up then we should do it, not wait for the mock ups to discuss.

Jane- wait for the current release and then bring it up with WG.


  • Amelia's item-

It's basically that I think we need more users in the AmiGO working group, and that we confine the AWG chats to talking about issues that directly affect the users and focus on issues rather than on admin issues. Comments from Plant community? We want to develop AmiGO for GO, not for Plant. We can still get other input. Keep these meetings very focussed is important. Also we need to keep the discussions connected.

Action Item- Eurie/Jane will bring up the issue of inviting non consortium members to the WG discussion in the next Managers meeting.