TR_02SEP09

Agenda

1. Introduction and Q&A about iPlant, iPTOL & the tree reconciliation WG (SM)
2. Introduction to the collaboration environment (SM)

• https://pods.iplantcollaborative.org/wiki/display/iptol

3. Preparation for Sept. software developers meeting (SM & TJV)
4. Logistics and preparations for a November joint NESCent-iPlant meeting (JL-M)
5. Recruitment of scientific personnel (TV)
6. Scheduling of agenda and participants for future calls (TV)

Notes

Attendees: Benjamin Vernot, Cecile Ane, Nicole Hopkins, Sheldon McKay, Todd Vision, Karla Gendler, Dannie Durand

1.    Introduction and Q&A about iPlant, iPTOL & the tree reconciliation WG (SM)

McKay gave an introduction of iPTOL and the working groups along with a brief introduction to iPlant.

Vision stated that the Tree Reconciliation working group is focusing on molecular biology/genomics audiences and that gene trees will be Tree Reconciliation’s charge. 

Durand asked if there is a separate goal that there be a large collection of gene trees?  Would certainly like to add to them, but remember that iPlant is not necessarily a data repository (has this changed b/c it hamstrings the group if can’t store data).  McKay is to follow up on this point and get further clarification from the executive team.

What exactly is the Data Assembly working group assembling?  Focusing on a few genes that are well sequenced across plants.  Yes, they are gene trees but just genes that are widely used across plant genetics, so maybe just 2 genes.
Vision then asked of the group what would be most valuable outcomes of the working group? 

Ane said that she could only restate what the goal of the working group is; one goal is to have deliverables to be used by the community in terms of tree reconciliation; the big species tree not delivered by this group but this group will provide gene trees and methods to reconcile on this scale along with expressing/displaying uncertainity.  Durand commented that she missed the fact that there wasn’t a repository of gene trees but she sees the goal as being able to write marquee papers that would demonstrate the utility of the tools; issue of scalability; raises fundamental algorithm problem; issue of what to do with polytomies of interest (lineage sorting); with such a large species tree, there will be issues of polytomies.  Vernot commented on there being a large collection of gene trees and having the ability to measure uncertaintiy with branch lenghts and bootsraps.  Durand pointed out that if data is heterogenous (values mean different things and come from different methods), limits ability to reconcile; how do you automate edges and tree rearranging (have bootstraps, posteriors, branch lengths)? 

Vision stated that we can’t get around not having a repository of gene trees; hoping can use iplant software engineering for data interchange; for marquee analysis, necessary to generate/process trees in some uniform way;  JLM’s goal is to gather data for construction of trees; would like this group to be responsible for deciding how to construct these trees using iplant hardware; but will need a repository (is there a long term commitment to hosting and curation?), there’s a way to do this with frozen data sets.  The repository may not be as dynamic as wanted but at least there would be a repository.  It is worth thinking about adding tree skeletons so that a dynamic system would be more computationally feasible; easier if skeleton is there; put investment to have quality there; adding additional data/genes to the quality is easier than de novo.  With having the same data sets, methods can be easily compared. 

If a person can see gene trees reconcilied with the big species trees, then this group needs to think about viz, where the data is coming from (data exchange), marquee analysis that shows value of having large scale reconciliation against large trees (need to assess scalability).  Gene tree reconciliation methods are still in their infancy (rooting, uncertainity, confidence values, appropriate objection functions) so part of ET role would assess accuracy and scalability of initial methods; large set of simulation studies ; biological questions for marquee: id gene content in ancestral extinct species; multi-domain evolution.  However, the multi-domain problem might be too ambition for a two-year project.  The working group needs to provide requirements to core about how to interact, exchange with other repositories and own data

2. Introduction to the collaboration environment (SM)
•    https://pods.iplantcollaborative.org/wiki/display/iptol
•    Preference is to have centralized collaborative document management

3. Preparation for Sept. software developers meeting (SM & TJV)
One of the desired goals of the developers meeting is to work through problem statements from TE and TR working groups to generate requirements.  Thus problem statements are needed from this group.  Problem statements should be high level and from the biologist perspective.  In consructing problem statements, it is necessary to think about what is needed; what is it that you are trying to get to, what are your needs; high level, this is what I am doing, this is what I want to be doing; don’t want to focus on solution, tools; just what is it that you are trying to accomplish and what is it that you can see are the limitiations.  Don’t limit creativity; this is brainstorming session; we’ll go back and limit and look at feasibility.  The group should start generating prose and iPlant will begin to look at it.

4. Logistics and preparations for a November joint NESCent-iPlant meeting (JL-M)
The point of meeting is to parcel out tasks to scale up to larger analyses; Nov 23-24, possibly in PHX; would like to see this wrapped into the tree reconcilation working group as will have access to HPC
Can they get remote participation?  (need to follow up)
What about travel expenses? Depending on inivter group, iPlant will cover expenses if iplant sponsored event

5. Recruitment of scientific personnel (TV)
Delayed a week

6. Scheduling of agenda and participants for future calls (TV)
Two-pager to discuss, recruitment of personnel, online document that begins to layout ideas/goals of group