DI_07OCT09

Agenda

Val to discuss experience in managing a large tree in the Crimson (http://kim.bio.upenn.edu/software/crimson.shtml)
Discussion on moving Data Integration WG meetings to a bi-weekly schedule (Michael)
Requesting high level plan for taxonomic intelligence (Sheldon)
Taxon markup prototype (Bill, Jerry, Sheldon)
Coordination with other working groups (Sheldon)
Other general tactical items

Sheldon McKay, Bill Piel, Val Tannen, Jerry Lu, Scott Menor, Liya Wang

Val talked experience in the Crimson.
It is a simulation component of the CIPRES. It manages a simulation tree with 1 million leaves and 2 million nodes.
It loads and keeps the topology of the tree with minimal connection in newick format in memory. It retreieves other meta data associated with nodes in relational database with node id as the key. This takes 30 to 40 Mb memory and solves the performance issue.
User can specify leaves to find least common ancestor and other operations.
The source code (in Java) and documentation are availabe.
ACTION ITEM: Sheldon will follow up with core software development team to to see whether they have interets in the Crimson.

All agreed on moving the WG meetings to a bi-weekly schedule starting from this week.

Sheldon wants some concretet plans to find solutions to the taxonomic intelligence.
ACTION ITEM: Sheldon and Bill will prepare some outlines on this for next WG meeting.

Bill, Sheldon and Jerry met last week and discussed building a nexml decorating service, which will parse user's nexml, populate it with markup meta data, and get the decorated nexml back to user.
Bill agrees to provide helps on the data logic.
ACTION ITEM: Bill, Sheldon and Jerry will follow up on building the prototye of service.

ACTION ITEM: Bill will attend other WG's meeting a week from next Monday. (Data Assembly, possible also Tree Evolution and Tree Reconciliation)