Past Awardee

Toward an 'In Silico Plant' Research Platform

Xinguang Zhu

College: Life Sciences
Award year: 2008-2009

Human society is facing the grand challenge of providing sufficient food and energy while simultaneously preserving our fragile ecosystem on this planet. Plants sit at the center of solving all these problems. They are the source of our food and energy, and they take up the major greenhouse gas (CO2) from the atmosphere. Our long term goal is to develop an in silico plant research platform that can accurately predict the plant performance and identify where to engineer plants for higher productivity under different environments. There are substantial challenges in developing both novel models but also cyber capacities to achieve this goal. Through this fellowship, we will collaborate with the cyberinfrastructure directorate of NCSA led by Dr. James Myers to tackle the major technological barriers to build the in silico plant research platform and implement the basics elements of the in silico plant research platform (hereafter in silico plant). Through this fellowship, we will implement the following basic elements of in silico plant. A) Cyber-enabled WIMOVAC (Humphries & Long, 1995, Zhu, Humphries & Long, 2001). WIMOVAC (Windows Intuitive Model of Vegetation responses to Atmospheric and Climate change) is a generic model of plant photosynthesis and related plant processes. We will wrap WIMOVAC and enable online simulation. B) Work flow management. Capitalizing NCSA's experience in developing MAEviz Earthquake Hazard Analysis environment, we will implement workflow management in the in silico plant platform, so that modeling, visualization, and analysis can be done carried seamlessly and easily. C) Connections to Databases. In silico plant will use the expertise of NCSA in data and provenance management to improve the connection of WIMOVAC with external database and support comparisons across runs done under various conditions. D) Application support tools. We have developed a series of basic modeling support tools, including optimization, sensitivity analysis tools and comparison of modeling results with experimental observations. E) An extensive help system and detailed training material. The success of this basic in silico plant research platform will open up a range of new research, education and grant application opportunities. For example, with the access to supercomputer resources, in silico plant will enable us to simulate plant productivity over millions of sites, each with different weather and soil conditions. Such capacities, with the future inclusion of GIS capacities and more detailed modules for plant, soil, and hydrological processes, can make in silico plant a decision support system for the biomass-based bioenergy production. Furthermore, once this platform is available, we can begin to link our detailed cellular metabolic models (Zhu, de Sturler & Long, 2007) with WIMOVAC, to enable, for the first time, simulation of the plant production system from molecular level all the way up to crop field performance and beyond. Such multi-scale modeling research is currently heavily invested in by several federal funding agencies, e.g. DOE, NSF (such as the iPlant initiative). In addition, the in silico plant research platform can also serve as a primary education tools for students from middle school up to graduate level.