LEGO August 23rd 2011
Jump to navigation
Jump to search
LEGO call
Tuesday 23rd August 2011
Present:
- Chris Mungall
- Jane Lomax
- Emily Dimmer
- Rebecca Foulger
- Tony Sawford
- Pascale Gaudet
- Paul Thomas
- Amelia Ireland
- Rama Balakrishnan
Aim of the call
- To come up with a broad plan of action on how to manage the LEGO project, what's in the scope of LEGO, what's out the scope of LEGO and how to proceed.
Relevant documents
Chris went through his proposal, which is designed to be read alongside Paul's White paper.
- http://wiki.geneontology.org/index.php/LEGO_Model_Draft_Specification
- http://wiki.geneontology.org/index.php/File:Paul%27s_LEGO_white_paper_March_2010.pdf
Summary of call
- Need to separate (i) the LEGO idea, and (ii) the implementation of LEGO. Therefore, after a brief discussion on annotation IDs and how they'd be maintained by smaller groups without much man-power, it was decided to leave the annotation IDs for a separate discussion later down the line.
- We'll have to come up with displays to hide the 'ugly/complicated' stuff from users. This will be covered with the implementation discussions.
- Right now, we need to take a step back and collect the use cases, justifications, what problems we're trying to solve with LEGO, WHY GO should be doing this, and why it's different to a pathway database (Chris and Paul have ideas on this).
- We'll always need a mixture of pre- and post-composition for annotations and terms, because there's no ideal level of either.
- AI: Add annotation examples where you'd like to capture additional information (e.g. timing of a process, targets of a process etc) to http://wiki.geneontology.org/index.php/LEGO-style_annotation_ideas
- Jane will add a virus example.
- Need to start capturing relationships, alongside the relationship work that's ongoing for column 16.
Other points
- We can't currently capture that a GP is a TARGET of a process, where you don't know the exact GP doing the process (e.g. that a protein contains a caspase cleavage site, but you don't know which caspase is cleaving it) (ECD). LEGO will be able to capture this using, for example, PRO IDs for a generic caspase (CJM)
- Need to make nested statements about which processes are involved in other processes (refer to the NEDD4, RNAPII and response to diagram in Pauls pdf). (PT)
Remaining Questions
- Can any of this information be captured in the current GAF format, so we can move forward with this without waiting for the common annotation framework (CAF)?