Cyberinfrastructure for Biology
The i Plant Collaborative

The last project I will discuss is the iPlant Collaborative (iPC), a cyberinfrastructure project very recently funded by the US National Science Foundation. A collaboration among the University of Arizona, Arizona State University, Cold Spring Harbor Laboratory, Universityurdue) and the Carolinaniversity of North), iPlant is receiving $50M over five years to create a cyberinfrastructure collaborative for the plant sciences that will enable “new conceptual advances through integrative, computational thinking.” The project began in February 2008, and so is still in its organizational stages.

In contrast to the technology-driven cyberinfrastructure projects discussed earlier, iPlant focuses more on the human side of the infrastructure than its technical side. Roughly a third of its budget will be devoted to community building via a series of symposia, workshops and meetings that bring together plant scientists, computer scientists, software engineers and mathematicians to identify and discuss grand challenge problems in plant biology, including such fundamental questions as how genetic diversity in plant populations translates into phenotypic diversity, and how plants perceive and respond to the environment. Participants of these meetings will be asked to identify ambitious but feasible research projects that address some of these grand challenge questions.

The iPC investigators expect that a small number of novel research projects will emerge from these discussions, and that some of the participants will team up to form collaborative research teams to take up some aspect of a grand challenge. iPC will provide these grand challenge teams with the basic infrastructure needed to support their collaborations, including physical meeting space, mailing lists, videoconferencing systems, web pages, WIKIs, blogs and electronic forums, as well as customized software resources which iPC calls “discovery environments.” These are envisioned to be grid-like collections of data and compute resources that have been organized in a way most suitable to the grand challenge team’s research needs. For example, a discovery environment geared to a developmental biology project might provide access to anatomy and developmental ontologies, histological image databases, annotated collections of developmental mutants, signaling pathway databases, and to simulation tools for modeling intercellular communications.

Although designed to meet the needs of a specific grand challenge team, the discovery environments are intended to be open to the whole research community. To maximize their generality, discovery environments will have extensive “mash-up,” facilities. The mash-up, a concept that should be familiar from Google Earth, allows disparate datasets to be tied together via a framework. The discovery environment mash-up facilities will encourage researchers to combine their own data sets with public datasets and with those of their colleagues in the hopes of discerning novel patterns that would otherwise be inapparent. For example, a developmental biology discovery environment might use an interactive diagram of signaling pathways as its mash-up framework. One part of the research time might superimpose on top of this framework a set of microarray-derived gene expression patterns from developmentally normal and mutant plants, while another could contribute a transcription factor-binding site interaction dataset taken from a series of ChIP on chip experiments. The combination of these data sets might point to a hypothesis for the mechanism of action of the mutant, which could then be explored using simulation and modeling tools attached to the discovery environment. The resulting model would be published back to the discovery environment for use by the rest of the team and the broader research community.

Unique among the projects discussed earlier, the iPlant Collaborative has sociologists on staff to monitor the effect the iPC has on patterns of interdisciplinary collaboration and to recommend ways to improve the interactions. It also has a significant education and public outreach component. Only time, of course, will tell whether this mixture of technical and human infrastructure will live up to its promises.