The complex metabolic networks of eukaryotic cells pose significant challenges for model reconstruction, with the compartmentalized nature of eukaryotic metabolism and limited understanding of metabolite inter-compartmental transport being primary blockades. Recently, constraint-based modeling (CBM) techniques have extended to studying multicellular organisms such as plants. This work describes the whole-plant genome-scale reconstruction of switchgrass (Panicum virgatum L.), a perennial grass native to North America and a promising bioenergy crop. The compartmentalized model of this C4 plant spans more than 9,000 genes and describes the interactions that occur between mesophyll and bundle sheath cells in the leaf during photosynthesis, as well as the flow of nutrients to the stem, roots and panicle through the phloem. We describe the reconstruction process of the lowland genotype) (i.e., AP13) model, including the use of gene expression data. The model maps metabolites and reactions into organ- and compartment-specific sub-models which provides a framework for accounting for genetic diversity. Application of the model captures diurnal cycling, identifies reactions required for growing and mature leaves, and finds changes in reaction flux variability profiles under simulated stress conditions. Our multi-year field data collected as part of this work informed time/condition-based constrains for model predictions. These measurements include both phenotype data (e.g., height, stem diameter, tiller count, fresh/dry weight, and weight of separated stems, leaves, panicles), and organ biomass composition variation in lignin, carbohydrate, and protein levels, with lignin ratios being found using Molecular Beam Mass Spectrometry (MBMS). We apply these constraints in a seasonal growth simulation framework to examine biomass production over three growing seasons and examine potentials for improving harvest yield.
