Plant modelling is speeding up crop improvement by testing outcomes in silico, but how can these models be made more effective? Some argue for more complexity to better model biological reality, while others say parsimony could lead to insights to the facts that are most important. Hammer and colleagues argue in a new paper in in silco Plants, that it should be possible to do both.
“Biological realism in crop modelling requires formalisms based on insights on ecophysiological mechanisms at plant/crop scale as well as on insights on metabolic processes at cellular scale,” the authors say in their paper. “Parsimony in crop modelling requires frugality of assumptions and detail in order to achieve robust predictions of crop growth and yield—as simple as possible but no simpler—across diverse genotypes and environments. Multiscale models that operate effectively across levels of biological organization provide an avenue for advance.”
Hammer and colleagues said that the multi-scale nature of the next generation of models could combine complexity with speed. “Models structured to readily utilize algorithms operating at varying levels of biological organization, while using coding and computational advances to facilitate high-speed simulation, could well provide the next generation of crop models needed to support and enhance advances in crop improvement technologies. Hierarchical algorithm nesting is a means to link approaches operating at differing levels of complexity and biological organization while retaining biological realism at all levels.”
The authors say that multi-scale models will not only benefit plant breeding, but also the scientists working on them. By working at multiple scales, the models have relevance to scientists working in different fields. In the paper Hammer and colleagues said: “Chew et al. noted that their integrative modelling operating at the interface of several research communities had the potential to facilitate communication and draw together the different types of understanding from fundamental plant research and crop models.”
“The need for effective transdisciplinary dialogue and connectivity is clear. Committed teams with shared vision and effective leadership targeting the building of cross-scale models with a clear purpose provide a means to achieve this.”