The genomic sequence of Escherichia coli appeared in 1997. Since 1998, this R01 program has systematically built genome-scale models of E. coli, culminating with models of Metabolism and protein Expression (ME models) that compute up to 80% of the proteome by mass in rapidly growing cells [1]. The models generated under this program have deepened our understanding of how to read genomes, computationally model the physiological processes they encode, and to guide design of interventions. These genome-scale models have enabled hundreds, possibly over a thousand, systems biology studies of bacteria. Half of the unmodeled proteome represents stress functions that provide responses to oxidative, thermal, and acid stresses. The human immune system uses these stresses to eradicate pathogens, but no mechanistic model can currently compute how these stresses perturb key cellular processes on a genome-scale. Therefore, new modeling methods are needed. Our preliminary data strongly indicate that the underlying molecular mechanisms can be modeled by extending the ME modeling approach. Our laboratory has also developed Adaptive Laboratory Evolution (ALE) technology capable of generating hundreds of evolved strains and high precision DNA assembly protocols to find all mutations occurring during ALE. We can thus computationally model, experimentally evolve, molecularly profile and mechanistically determine the genetic basis of stress tolerance. We propose an iterative three-step process that will 1) EVOLVE E. coli under various stress conditions, 2) ANALYZE the resulting phenotypes through data analytics and mechanistic modeling, and 3) VALIDATE model- driven hypotheses. This iterative workflow takes advantage of our ALE technology, extensive mutational and RNA-Seq databases with accompanying data analytics, and genome-scale modeling capabilities to elucidate cellular responses to oxidative, thermal, and acid stresses. A major outcome of the proposed program is an experimentally-validated genome-scale model that employs novel methodologies to describe stress responses, metabolism and protein expression (called the StressME model) that increase computational coverage of the E. coli proteome up to 90% of proteome mass. In particular, the three stresses that we propose to study are critical for a deep systems-level understanding of the tolerance that pathogens have against the stresses imposed by the immune system and certain types of antimicrobials. The genetic basis revealed can be compared to characteristics of wild type strains isolated from patients. Thus, these new models will facilitate future translational studies that investigate infectious disease.