Developing the nuts, bolts, theoretical frameworks and community infrastructures to support global plant systems biology research

The rate of progress in biological science today can be compared with that of physical science at the beginning of the 20th century. With so many new discoveries emerging on a daily basis, are these discoveries simply the application or manifestation of basic scientific principles such as the central dogma of molecular biology, the fundamental theorems of genetics, or the basic genetic principles of epigenetics, or will they have a larger impact? Are there grand challenges to be found in biological science research beyond the current stage of fact observation and the elucidation of the underlying mechanistic basis of individual cases using the basic principles discovered long ago? The answer is ‘yes’. In the field of plant science, though there is ever-increased resolution of the mechanistic details of the plant growth and development, we are far from being able to predict the growth and development of plants of particular genotypes under different environments, let alone predict changes in plant growth and development for an altered genotype and under a changed environment. This is one major reason why we label biology as an ‘experimental’ science, which is essentially another way of saying that models for whole plants are far from being mechanistic, sufficiently robust and satisfactorily predictive. Is there any hope of developing such models given the complexity of biological systems? There are good reasons to be optimistic. Pioneering researchers have developed many models that enable the accurate quantitative prediction of biological performances, such as the prediction of photosynthetic CO2 uptake rates under either steadystate (Farquhar et al. 1980) or dynamic conditions (Zhu et al. 2013). Models simulating many other physiological plant processes have also been developed (see reviews in Zhu et al. 2015; Chang and Zhu 2017). It is foreseeable that, as modules for individual processes become available, robust and complete models for plant growth and development will be within reach. This will not be a simple task. Even for the model organism Arabidopsis thaliana, there are about 30 000 genes, whose action and interaction among them and with their micro-environments, underlie growth and develop. Creating highly robust models to predict the behaviour of such a complex system with detailed descriptions of the mechanisms of all underlying genetic, biochemical, biophysical, and associated physical and chemical processes represent a huge challenge ahead. If such models are developed, they can be used to support plant science research, such as to study the mechanistic basis of natural variations of plant structure and function, to study the responses, acclimation, adaption and evolutionary trajectories of plants under environmental changes, in addition to the currently widely appreciated roles of plant models in guiding crop engineering, breeding or cultivation (Zhu et al. 2015; Marshall-Colon et al. 2017; Xiao et al. 2017). There are a few research areas critical to tackling this grand challenge. First, identification of new biological entities or new regulatory mechanisms that control different aspects of plant growth, development and plant–environment interactions, will continue to be the area that produces great discoveries in plant science. Here I emphasize that, in addition to the standard qualitative characterization of individual genes and their products, such as the structure and regulation of mRNA and proteins, their dynamic or kinetic properties, such as their rates of synthesis, degradation and Michaelis–Menten constants should be another major focus of future research. These quantitative data are required for the development of rate equations describing the dynamic changes of these biological entities. Second, development of accurate rate