A modelling portal for the UK plant systems biology community

This project is all about plant growth models. Here, the word 'model' does not refer to a physical object but instead means a mathematical way of representing something. This kind of model consists of a series of mathematical equations which can be solved, normally using a computer. By feeding these equations with values for factors like temperature, amount of nitrogen available and light energy that plants can capture, the models make it possible to predict how a plant will grow. Many of the models available are for crop plants, and they can give estimates of how much a crop will produce, whether it is a wheat plant yielding grain, a tree growing new wood, or a pasture providing grass to feed animals. Some of the more complicated ones also represent interactions between plants and other species, such as bacteria and fungi which improve or reduce nutrient uptake by plant roots, or leaf-eating herbivores. Historically, these models work at the level of a whole plant, a farmer's field or even a landscape, because it has been most straightforward to collect data at these levels. Modelling plant growth, an area of research which has flourished for several decades, is one in which the UK has been especially strong, but which is not well known among modern molecular scientists. In the new era of 'systems biology' research, scientists are able to capture large amounts of data about processes that happen on a much finer scale - an individual leaf or root, for example, or even within a single cell. The challenge is to use this mass of information to predict how a plant's genetic makeup controls the way it grows and interacts with its environment. It is becoming clear that many of the mathematical techniques needed to make this possible are the same as those used by the crop modellers. The aim of the present project is to make existing crop models available to systems biology researchers via the Internet in a user-friendly way that is independent of any particular computer software. This means that systems biologists can adapt models for their own use, and connect them with other models that work at the cell level. We want to develop a portal, a one-stop Internet shop for the models themselves, examples of how they have been applied, and explanations detailed enough to allow other people to adopt them. We will also provide a forum so that the community can add comments and suggestions for further development, and help each other in using these and other models in the future. In addition to plant systems biology experts, the portal will be of use to crop scientists and to policy-makers. Predicted climate change will mean that our current crops will perform differently in the future - the models will help predict whether the difference will be beneficial or cause loss of yield or failure of crops. They will also aid researchers and breeders in developing crops for new uses, such as bioenergy. Some models have the capability to predict the behaviour of whole ecosystems, such as forests, under changing conditions. Models are important for the future of plant science research, agriculture and the environment and it is important to make them as widely accessible and usable as possible / this is the purpose of our project.

The project will develop a portal to make available, in open-source format, plant growth models relevant to the needs of the UK systems biology and crop science communities. A Wiki will allow researchers from all fields to provide input on the models. Seven existing models have already been selected for implementation and at least 3 others will be chosen through discussion with the community. Models available in a range of formats and programming languages will be represented in an open-standard declarative XML format, using an established modelling environment. A declarative language allows models to be displayed in various ways and efficiently makes model metadata available automatically. The XML generated for each model can be run through one of several code-generators (some already implemented), allowing users flexibility in the language and computational platform used for implementation. We will provide annotated examples of typical output for each model and devise simulations and analysis for publication and dissemination on the portal. Demonstration results will use the high-level models developed here and also models at other scales. The portal will integrate the models with software tools used by systems biologists and will make models available to users at external sites, via the Systems Biology Software Infrastructure at the Centre for Systems Biology Edinburgh. To extend the Systems Biology Markup Language to represent plant growth models not expressible in biochemical-pathway terms we will collaborate with Dr N le Novère (European Bioinformatics Institute), who is responsible for the BioModels portal. Further links between the project participants and the SBML community, other UK Centres for Integrated Systems Biology and EBI will keep the portal's development abreast of international standards development, including plant and trait ontologies, and allow the project staff to provide input to the development of later model exchange standards.