CleanUpMeeting09Agenda/
Contents |
Agenda for October 2009 Cleanup meeting
Agree a process for OBI 1.0 release, clean up issues below for discussion:
Meeting info:
The Cleanup workshop will be held in the 13th floor library of Blockley Hall. 423 Guardian Drive, Philadelphia, PA, 19104 except for Sunday, the first day, where we will meet in the 14th floor conference room. These are in the Penn Center for Bioinformatics.
Parking is available in a secured parking garage at 38th & Walnut ($20 per day) and an open lot just pass the University Avenue bridge ($7 per day I think).
See http://www.cbil.upenn.edu/meetings/2009-10-15_BlockleyParking.png
Hotel info:
If you plan to attend the clean up meeting in Philadelphia and need a place to stay, you may find difficulty in booking a room especially on Sat. Oct. 17 (our meeting is scheduled from Sunday Oct. 18 to Wed, Oct. 21). Not sure what's going on - perhaps the Philadelphia Film Festival?
Anyway, when I checked yesterday there were still rooms available at hotels near the airport on Sat. Oct. 17. There is a Marriott ( http://www.marriott.com/hotels/travel/phlar-philadelphia-airport-marriott/ ) attached to the airport but others close by. (try Google maps with "hotels near philadelphia airport")
To get into town, there is a train, the R1 that runs from the airport into Philly with a stop at Penn. http://www.septa.com/service/rr_schedules.html However, I can also give people rides as well.
For the rest of the dates there are B & Bs such as http://www.gablesbb.com/reservations.html Here's a site with other B & Bs. http://www.bedandbreakfast.com/philadelphia-pennsylvania.html
If you are feeling lucky, the two hotels nearest Penn are the Sheraton (http://www.philadelphiasheraton.com/) and the Inn at Penn (http://www1.hilton.com/en_US/hi/hotel/PHLIDHH-The-Inn-at-Penn-A-Hilton-Hotel-Pennsylvania/index.do). but these were booked when we checked.
CS
Meeting Agenda
Sunday 18 October
- Manuscript Discussion - all to have reviewed prior to meeting and a new version will be available
- Examine Tracker Items, progress, prioritize - add more if needed
- Process, we will operate a cut off on discussions, so if we can't make a decision we will move on. Need to agree what this will be.
P9 Items
- 2845352 ISA-tools use case
- 2843699 person/Homo sapiens in OBI
- 2843677 IAO section of paper
- 2843658 use of plurals in the manuscript
- 2832693 Clinical study use case
- 2100531 mixtures of chemicals in Chebi
P8 Items (need prioritizing)
- 2877331 helper classes related to device
- 2875994 reagent roles
- 2872993 change import order of use case
- 2855265 instument clean up
- 2836360 Review all necessary and sufficient conditions for accuracy
- 2828871 _cloning placeholder
- 2613561 action specification
- 2570459 Device setting
- 2078795 vector role (cloning and disease)
- 1889390 Objective vs. Objective-neutral Planned Processes
Paper Related P8
- 2832725 Supplementary Table of terms used in manuscript
- 2832701 Update of figure 1 in manuscript
- 2845347 Authorship, Affiliations, Funding
Release Related P8
- 2843660 calculate the no of classes etc for release
- 2878327 curation status of OBI terms
Monday 19 October
Tuesday 20 October
Wednesday 21 October
Thursday 22 October
General clean up items
Remove _classname - placeholder classes
e.g. _realizable entity, _under discussion
note: _realizable entity and other top level buckets are only there to provide an anchor for defined classes. I would vote in favor of keeping those "development helper" classes in a separate file, along with OWL2 bits and all things we don't want in the released file. Other example of unwanted things is disjoint between obsolete classes and bfo:Entity, as this may create inconsistencies for users importing OBI and somehow defeats the purpose of obsoleting classes.
Remove editor notes
some difference of opinion here, BP/HP want to remove these, AR thinks they add value MC: happy to distribute editor notes if those were cleaned up. I don't think that this will happen realistically, nor that it should be a priority before release, so I would vote for removing them.
Resolve all obsolete term useage
should be done by next release, will need checking again
Resolve classified tree wierdness
probably needs a manual check of defined classes
Remove duplicates
e.g mass spec
Decide what to do with issues relating to requests to modify other ontologies e.g. RO
resolve/workaround/leave?
Check leaf nodes with a single child, and check there potential additional children
Clean up quality branch (aka conferred/historical qualities)
Release related items
Use of obsolete terms
Extracted from the deprecation policy google doc - Implement scripts to find obsolete classes by either: deprecatedClass/deprecatedProperty, "obsolete" curation status, or being a subclass (direct or inferred) of ObsoleteClass or ObsoleteProperty. Update all such cases to remove axioms other than the single subclassOf, to update the curation status, and, optionally, to mark as owl:deprecated. tracker item
Resolve editor attribution
Check inconsistencies
Produce a saved inferred version
Mireot terms
anything mireoted should probably be up to date in OBI 1.0 - do we have means to check this?
Minor issues to resolve now
Enforce curation status
Two issues
- 3.1. curation status is not up to date and needs to be for release
- 3.2 review terms where curation status is really not complete and prioritise if any need to be done for release, and add to tracker
Spellcheck, check for consistent class naming style
Resolve cases where there are no logical/text definitions
Resolve inconsistent logical/text definitions
Clean up editor attribution policy is one editor's name per annotation property, using PERSON: xxxxxx
