This paper presents an expanded genome-scale metabolic model of Escherichia coli K-12 (iJR904 GSM/GPR), which includes 904 genes and 931 unique biochemical reactions. The model is more comprehensive and chemically accurate than the previous model (iJE660a GSM), which included 660 genes and 627 reactions. The new model incorporates gene-protein-reaction (GPR) associations, allowing for direct integration of diverse datasets such as genomic, transcriptomic, proteomic, and fluxomic data. The model also includes elementally and charge balanced reactions, and has been used to identify putative assignments for 55 open reading frames (ORFs). The model has been validated against experimental data and has shown improved predictive capabilities, particularly in scenarios involving proton exchange and growth on different carbon sources. The model also accounts for the specificity of quinones in the electron transport chain, which was not considered in the previous model. The model has been used to analyze the effects of gene deletions and to identify metabolic gaps in the network. The results show that the new model provides more accurate predictions of metabolic processes and can be used to better understand the genotype-phenotype relationship in E. coli K-12. The model also highlights the challenges of globally balancing protons in cellular metabolism. The paper also discusses the use of phenotypic phase planes to compare the metabolic behavior of E. coli under different growth conditions. The results show that the new model provides more accurate predictions of optimal growth conditions and metabolic flux distributions. The model has been used to identify new metabolic routes that enable anaerobic growth on α-ketoglutarate. The model also includes a detailed description of the metabolic network and the reactions involved in the TCA cycle and citrate lyase. The model has been validated against experimental data and has shown improved predictive capabilities. The paper concludes that the new model provides a more accurate and comprehensive description of E. coli metabolism and can be used to better understand the genotype-phenotype relationship in the organism.This paper presents an expanded genome-scale metabolic model of Escherichia coli K-12 (iJR904 GSM/GPR), which includes 904 genes and 931 unique biochemical reactions. The model is more comprehensive and chemically accurate than the previous model (iJE660a GSM), which included 660 genes and 627 reactions. The new model incorporates gene-protein-reaction (GPR) associations, allowing for direct integration of diverse datasets such as genomic, transcriptomic, proteomic, and fluxomic data. The model also includes elementally and charge balanced reactions, and has been used to identify putative assignments for 55 open reading frames (ORFs). The model has been validated against experimental data and has shown improved predictive capabilities, particularly in scenarios involving proton exchange and growth on different carbon sources. The model also accounts for the specificity of quinones in the electron transport chain, which was not considered in the previous model. The model has been used to analyze the effects of gene deletions and to identify metabolic gaps in the network. The results show that the new model provides more accurate predictions of metabolic processes and can be used to better understand the genotype-phenotype relationship in E. coli K-12. The model also highlights the challenges of globally balancing protons in cellular metabolism. The paper also discusses the use of phenotypic phase planes to compare the metabolic behavior of E. coli under different growth conditions. The results show that the new model provides more accurate predictions of optimal growth conditions and metabolic flux distributions. The model has been used to identify new metabolic routes that enable anaerobic growth on α-ketoglutarate. The model also includes a detailed description of the metabolic network and the reactions involved in the TCA cycle and citrate lyase. The model has been validated against experimental data and has shown improved predictive capabilities. The paper concludes that the new model provides a more accurate and comprehensive description of E. coli metabolism and can be used to better understand the genotype-phenotype relationship in the organism.