TR_9SEP09

Agenda

1. Goals document

2. Problem statements for the iPlant developer meeting, Sept 2009 in Texas

3. Hiring

4. Nov 23-24 meeting logistics

5. Sankoff iPAT project

Notes

Attendees: Jerry Lu, Ann Stapleton, Sheldon McKay, Dannie Durand, Todd Vision, Ben Vernot, Jim Leebans-Mack, Cecile Ane

1.Goals document

• Vision presented the goals document, wanting to make sure everyone is in agreement with the overarching goals and also milestones. 

• The working group needs to look at what’s feasible, useful, off target, necessary and then begin to refine the project plan based on this.  This working group is an appendage of the overall project. 

• Both Durand and Vernot agree with the overarching goals.  Vision commented that a big task of this working group is going to be throwing out a lot of good ideas because they don’t match the overall goal.

• In discussions relating to the subgoals oulined in the goals document, Vision asked what previous work has been done to deal with performance/scalability/uncertainity that hasn’t been published. 

• Durand answered that with NOTUNG that really haven't hit problems or scalability issues yet. 

• Vernot asked what specific problems will be tackled by this group.  With NOTUNG, can do things with nonbinary species or nonbinary gene trees but not both (supposedly NP-hard); have heuristic for it; haven’t hit scalability but will probably will run into it.  Vision is curious what happens as both the gene tree and species trees get larger. 

• Durand and Vernot agreed to put a short, "hand-wavey" summary on the wiki about what they have seen with scalability with NOTUNG.

• Another issue that needs to be address is accuracy; how does performance degrade when topological error and too much uncertainty?

• There is also lineage sorting at tips versus anywhere but Vision stated that lineage sorting is much more of a problem at the tips.

• Durand asked if there is a formal definition of accuracy in context of tree reconciliation? 

• Vision said there are three measures that this group needs to look into: accuracy, uncertainty (way quantifiable and inferable) and reconciliation similarity score. 

• Looking at the subgoals, Vision sees 1-3 as review papers with methodologies, 4 as implementation and 5 as the marquee analysis.  Durand suggested that points 1-3 are not entirely review papers; they could be theory papers, especially if no one has thought about the three measurements. 

• Durand asked if we wanted to add anything about algorithm development to the list of subgoals.  The existing algorithms are there because constrained by time; algorithm research would be good if we have the expertise but that this group can't just be forward looking as it will miss the opportunity to make progress now. 

• Durand states that her lab has been thinking about things so perhaps they can be implemented.

• Vision has asked that the group keep adding to the list of subgoals and that we can then decide what points we want to address in the next year;

• As for visualization, McKay said there is considerable overlap between TR and TE and TV working groups. 

• Vision countered with the hard part for viz will be with creative part; what will the visualization need to look like and this will need to come from this working group.  McKay will get back to Vision on who is on the Tree Viz working goup.

2. Problem statements for the iPlant developer meeting, Sept 2009 in Texas
Moved discussion offline for time reasons; Nicole Hopkins to pick up discussion; very informal user scenarios;

3. Hiring
Vision has received CV from Ben Vernot and has asked for CV from John Bowers who is a post post-doc at Georgia.  The skills complement each other; John is less of a programmer, writes pseudo-code; Ben is much more of a CS;

• Ben is time limited 50% until next summer; like to get a percent effort and time line for each of them.  John would be available 50%, starting ~November.  With Ben it depends on the extent of the overlap between what he is doing now and what needs to be done, if issues of could start immediately.

• Leebans-Mack suggested that it would be good to have a good outline of what responsibilities are in advance for each of the positions. 

• In regards to the splitting salaries and the difference between the iPlant amount and what they actually make, differences could be made up by their respective home labs.

• Durand commented that as we develop subgoals, could then look at how the person would fit in and how much time would they actually need.  The SOW shoudl be worked out as a group and then given to the respective business offices. 

• Vision stated that there is still summer salary but people will need to be able to commit time to get the money.  For Durand, she would like it spread over the year; for Leebans-Mack would have to be summer salary; Ane needs 1.8 months of salary.

• As of now, Ben would have a start date of October 1st and John start date of November 1st.

AI: send CVs of each applicant to the group
AI: SOWs need to be developed offline

4. Nov 23-24 meeting logistics

• For the particpants that are overlapping with the iPToL workshop, it is necessary to send the invitations right away.  Everyone from the WG will need to be added also. 

• Within the next day or 2, JLM will provide McKay with the invitee list.. 

• From a logistics standpoint, we might have to move the iPToL kickoff meeting attendees to Sunday afternoon. 

• McKay will look into remote particpation hosting by iPlant.

5. Sankoff iPAT project

• Stapleton presented the iPAT project and asked if it is a tool that fits with TR and if this tool is useful and in what way would it be useful.  Durand stated that it is a useful tool and that looking for conserved blocks, or looking at how things are happening spatially would sit very well with the overall them of the this working group.  JLM agreed and added that Bowers would like to add chromosomal location to reconcialtion process, as both an input and a way to look at information at the end.  Stapleton added that from logistic point of view that she will put grad students and post docs in touch with Nicole Hopkins to get reading material on gathering/writing requirements; students will have to revise charter and then get back to the group; people who have never done this before so there will be a lot of leg work up front; might be worthwhile to have them look at the wiki.