iPlant Collaborative

There are several barriers to addressing ‘grand challenges’ in biology: Biologists lack the ability to make complex queries across heterogeneous data sets that are usually incomplete and arbitrarily organized. Groups develop their own file formats, data models and application software without awareness of comparable needs elsewhere. Thus, it is almost impossible to combine data and models from multiple sources and utilize the most up-to-date tools to analyze and simulate complex inter-relations. More globally, computational thinking and collaboration is diffuse because the goals of computational scientists and biologists are often not well aligned. An effective cyberinfrastructure will encourage communication among disciplines and reuse of data models, file formats, application software and algorithms, while fostering the type of cross-disciplinary exchange of ideas that will advance both the specific case of plant science, and the broader fields of life and computer sciences. To address these challenges we are developing The iPlant Collaborative (iPC), a community of plant biologists, computer scientists, mathematicians and engineers organized around a core cyberinfrastructure.

An organizing principle of the iPC will be Grand Challenge Teams, cross-disciplinary, community-driven research groups which work collaboratively with iPC staff to design and develop ‘Discovery Environments’, software platforms custom designed to help the GC team address the selected Grand Challenge. Discovery Environments will typically take the form of “mashup” applications, which facilitate the integration of diverse types of data and tools; but beneath their surface simplicity, Discovery Environments will support sophisticated systems for semantic integration, description, and manipulation of biological data types. Discovery Environments will be integrated into the growing iPlant cyberinfrastructure, becoming in time an open source resource that is expanded and maintained by the community as a whole. Importantly, decisions for the allocation of iPC resources to specific Grand Challenges will be made a community-representative Board of Directors (BoD), not by iPC personnel. The BoD itself will be chosen by a “bootstrap” process in which a representative Nominating Committee will be selected by the PIs with NSF oversight to then select the Board of Directors. The role of iPC personnel is to facilitate community discussions and decisions and to build the cyberinfrastructure necessary for solving Grand Challenges. The BoD will make all major decisions about allocation of Collaborative resources to Grand Challenges prioritized by the community.

To initiate community discussion of the nature and tractability of Grand Challenges in plant sciences and the computational and cyberinfrastructure challenges that need to be addressed in order to solve such Grand Challenges, a “kick-off” conference on ‘Grand Challenge Problems in Plant Biology’ will be held April 7-10, 2008, in Grace Auditorium at Cold Spring Harbor Lab with several hundread attendees spanning all relevant disciplines, which will be publicly webcast to foster the broadest possible community participation. A balance of invited speakers, discussion time, and break-out group discussions will be scheduled. Major focus areas of the kick-off conference will include:

1. What is a grand challenge and what are the most important ones in plant biology? Presentations by leaders in the plant biology and CISE communities.
2. What are the data management, mining, discovery environment, and other infrastructural needs? Presentations by members of the Integrated Solutions and Infrastructure teams, as well as members of the community.
3. Training the next generation. Presentations by members of the community (including project team leaders) on “teaching computational thinking in biology” and how to use the iPC to facilitate this.
4. Invitation of proposals for Grand Challenge Workshops and Teams.

Next, the major focus of the iPlant Collaborative will be a series of small, focused, retreat-like Grand Challenge Workshops held at the Univ. of Arizona’s Biosphere 2 conference facility in Oracle, AZ, (and potentially also at CSHL’s Banbury Center). Each workshop will consist of 30-40 invited leaders from experimental and theoretical plant biology, computer scientists, information systems specialists, statisticians and engineers, and will be focused on areas of plant biology, e.g., plant development, plant-microbe interactions, ecological genetics, etc. These workshops will be designed specifically to encourage participants to go beyond their standard, canned talks to challenge experimentalists to enunciate the types of information resources and theoretical models they need to interpret the biological system under discussion. Theorists will be challenged to list the datasets they would need in order to create working models. In addition to community researchers, symposia will be attended by representative members of the Integrated Solutions, Infrastructure Development, and Synthesis Activities teams. Typically, the first half day of a workshop might consist of a meeting with the Integrated Solutions Team to learn about the iPC’s cyberinfrastructure capabilities and plans. This will be followed by a 2-3-day intensive workshop to address Grand Challenges in the subject area of the retreat. A final half day might be used for a follow-up meeting with the Integrated Solutions team to discuss the data analyses and cyberinfrastructure needs that have arisen during the retreat. (For retreats held at the Banbury Center in New York, meetings with the teams in Arizona will be held via web-conferencing, and follow-up visits to UA will be scheduled as needed). We anticipate that 6 workshops will be held during between August and December, 2008, and up to four in each subsequent year. From some of these GC workshops will arise the seeds of self-forming Grand Challenge Teams, who will propose the dedication of Collaborative resources to the development of Discovery Environments directed at solving the team’s Grand Challenge. The BoD will prioritize allocation of Collaborative resources to Grand Challenges.

We would anticipate at least two types of Grand Challenge symposia: one organized to assess the nature of grand challenges in a given discipline (such as plant-microbe interactions, signal transduction networks, or plant growth and development) and another that is driven by a self-forming group with an interest in a specific Grand Challenge. Proposals of each type will be evaluated with the assistance of a nationally representative group of reviewers/advisors with oversight by Board of Directors. Workshop teams will be configured to be diverse and multi-disciplinary. Workshops may lead to one or more proposals to create Grand Challenge teams, again to be vetted by representative reviewers and the BoD. Selected Grand Challenge Teams will come to UA’s BIO5 Institute to work with iPC staff as needed to plan and to develop a custom Discovery Environment for the project. UA faculty liasons will be assigned as needed to provide expertise and necessary local advising, mentoring, and oversight of students and postdocs working with Grand Challenge Teams and coordination with iPC staff. Visiting scientists may work in residence at BIO5 or CSHL for extended periods of time. To the extent possible, “common denominators” among Grand Challenge projects will be identified in order to help prioritize efforts. In addition, we anticipate that workshops may be held that focus on problems in biological data management and algorithms.