SUMMARY Metabolic enzyme velocity (the product of activity and abundance) is shaped by the need of growing cells to maintain metabolic flux while avoiding the accumulation of toxic intermediates and limiting inefficient biosynthesis. Imbalances in enzyme velocity play a significant role in metabolic disease, and changes in enzyme expression are often associated with drug resistance. Nonetheless, the connection between metabolic enzyme expression, intracellular abundance, and cell growth rate is poorly characterized. Our long term goal is to create a genome-scale model that quantitatively relates E. coli gene expression to growth rate phenotypes across both common laboratory and host-associated conditions. We envision using this model for antibiotic discovery, biosynthetic pathway engineering, and to interpret the effect of mutations in clinical isolates. Recently, we developed a new modeling approach that predicts the growth rate effects of combinatorial variation in E. coli gene expression and environment from sparsely sampled experimental training data. The basic strategy is to first quantify the growth rate effects of gradated changes in gene expression across multiple environments and genetic backgrounds of interest. Then, we use these data to parameterize a machine-learning model describing the connection between gene expression and growth rate. We found that the model can predict the effects of at least four combinatorial perturbations in gene expression and environment when trained on experimental data considering only pairwise perturbations. The central goal of this grant is to now apply and extend this approach to quantitatively understand the connection between enzyme abundance, growth rate, and antibiotic resistance in E. coli. More specifically, we will: (1) identify and model the metabolic factors influencing trimethoprim resistance in E. coli, (2) quantify the stoichiometric constraints on relative enzyme expression and abundance in 16 central metabolic pathways, and (3) apply new sequencing-based tools to simultaneously quantify gene knockdown effects and growth rate across core metabolism. Together, this work will test hypotheses about the modular organization of metabolism and yield a deeper understanding of the connection between metabolism and antibiotic resistance. Moreover, our work in Aim 3 will generate a genome-scale collection of barcoded strains for interrogating the dynamic connection between gene expression and growth rate across environments.