iPg2p Steering Committee Minutes
November 18, 2009; 1 to 2 pm CST
Present: Matt Vaughn, Steve Welch, Ed Buckler, Karla Gendler, Ruth Grene, Dan Kliebenstein, Chris Myers, Jeff White
The meeting was convened at 1pm CST and there was some general discussion to allow time for late arrivals to dial in.
Item 1: Review Action Items
- White post in-house training material that he has collected and solicit comments/additions
- Ann Stapleton has sent an example of training material and White asked what the next steps were? Welch asked if there is a need from other groups for a mutual understanding meeting much like the Visual Analytics group had. White thinks the modeling group will need a face-to-face.
- Gendler provide UA and CSHL press contact information to Welch
- Welch to check with Buckler about his press contact information
- Welch to draft the common text for the press release
- Welch has establish contact with Ag Press office at KSU and they are drafting a press release early this week. He will go over it and the press officer will make contacts with the press people that Welch has provided her.
- Goff to arrange a conference call between modeling group (White, Myers) and lignin group
- Gendler will follow up with Goff on scheduling this call.
- All provide feedback about data providers to Jordan, either on wiki or in email
- Gendler to query modeling group about setting up a modeling telecon
Item 2: Project Meeting Status
Project update meeting will be held January 13-14th, 2009 at the Hilton San Diego Mission Valley. Gendler needs to know who needs rooms, for what nights, and any dietary restrictions and will send out an email.
Item 3: Working Group Update
- postpone for next meeting
- Kliebenstein said the group has identified six potential users randing from the statistician/mathematician to a high school student. They have identified 3-5 large data sets for genotype-phenotype linkages. The group has also prioritized the statistical approaches: 1) GLM, to be used with GPU and parallelized; 2) mixed model; 3) Machine learning/Bayesian approaches. A framework is also being developed for creating simulated data sets. White commented on the environment characterization of NAM lines and crop management and stated that more data will be needed. Kliebenstein said this is a data integration issue thought GLM is being developed to handle envrionment. Welch added tissue type also. Buckler said for NAM lines can pull environmental data from the ag centers such as: temperature, rainfall, soil moisture, solar radiation, planting dates, fertilizer treament may be missing. It was not all dry land planting and he's not certain if irrigation records exist for those that were not but the data may be pullable. The group is also beginning discussions with the Viz group as the question remains how to view 25 million points and make it accessible to the reserach community. Grene commented that Greg Abram and Bernice Rogowitz might be able to help with computationally intensive visual needs.
- Grene stated the group just had a very successful Mutual Understanding meeting at TACC. The group wants tools that will integrate results from many different experiments. The idea is to propose workflows that can be visualized. An example workflow program is VizTrails which allows one to control inputs and outputs and also track provenance. Rogowitz and Brutnell have crafted a "Maize Genes Analysis" workflow and Usadel and Grene are working on creating a workflow for Grene's use case. After Thanksgiving, all will get a flow diagram of how the viz tool will work. Grene specifically wants a mechanism by which one can invoke any -omics data. The mutual understanding meeting was good because the plant biologists walked away with realistic expectations. Welch added that Provart's tool (eFP browser) will do 3D images of plant parts but is static. They talked about L-systems as he is using Google Sketch-up now. There is an opportunity to connect to the L-systems group. White asked if Python was being used exclusively and what is the strategy to avoid one-off solutions ot meet the use cases. Welch answered saying that the back-ends of the tools are very generic and there is interest in making the tools dynamic and more general. He also added that there is benefit to having development effort that will look into removing cumbersome-ness of the tools as they are but needs Dan Stanzione's input on if there is a move to a standard shell and how will people know that something is iPlant.
Item 4: Identify Action Items
- All: email Gendler with room requirements for Project Update Meeting.
Next meeting time: Wednesday, December 2, 1pm CST
Adjournment: The meeting ended at 1:40pm CST