The key evidence underlying the top-down construction, propagation, and analysis of genome-scale plant metabolic models

There is a growing demand for the effective metabolic and genetic manipulation of plants to benefit human society in many ways. The layers of complex interacting biological networks affecting plant growth makes this a difficult task, and several genome-scale metabolic reconstructions have been published as a means to enable researchers to explore these networks in silico. However the construction of these networks rely heavily on the evidence available for the biochemistry, gene-reaction associations, and compartmentalization. Furthermore propagating these networks for any species other than the model plant Arabidopsis thaliana in turn rely heavily on the phylogeny of plant protein families. I present the PlantSEED project where we set a gold standard for the curation of enzymes involved in plant primary metabolism, including isofunctional paralogs, from which we generate a draft top-down model of Arabidopsis thaliana and, via conservative use of EnsemblCompara families, 9 other species. I will also discuss the use of additional evidence to build a larger, compartmentalized model of Maize, and the use of transcript profiles to build. The data, along with additional functionality for metabolic flux analysis, is available in both the DOE Systems Biology Knowledgebase and our PlantSEED web portal. Finally I will describe our future plans for the expansion of PlantSEED.