||Abstract: "Genome-scale models (GEMs) of metabolism were constructed for 55 fully sequenced Escherichia coli and Shigella strains." P.20343 right column bottom paragraph: "Materials and Methods: The strain-specific model reconstruction procedure, gap filling algorithms, and in silico growth simulation conditions are described in SI Materials and Methods. Heatmap and phylogenetic tree and decision tree construction are described in SI Materials and Methods Carbon source and growth testing protocols are described in SI Materials and Methods. Eleven strains of E. coli and one strain of S. flexneri were tested as part of this study. The 11 E. coli strains are SMS 3–5, CFT073, HS, DH1, UMN 026, K011, Sakai, ATCC 8739, 042, EDL933, and K-12 MG1655. The S. flexneri strain was 2457T. E. coli 042 was a gift from Ian Henderson (Birmingham University, Birmingham, UK). All other strains were purchased from ATCC."
||P.20339 right column bottom paragraph: "Next, the models of strains known to reside intestinally were compared to investigate differences between commensal and InPec [intestinal pathogenic E. coli] strains. Models of InPec strains displayed an advantage in their ability to support growth on seven unique carbon and nitrogen sources (Table 2). Some of the substrates had unique enrichment specifically among the enterohemmorhagic (EHEC) strains, responsible for worldwide cases of diarrhea and hemolytic uremic syndrome (HUS) (ref 26)." P.20340 left column bottom paragraph: "In contrast to InPec strains, commensal strains displayed an advantage in their ability to degrade 11 unique carbon and nitrogen sources (Table 2). The short chain fatty acids (SCFAs) acetoacetate and butyrate were found to support growth as sole carbon sources for 78% of commensal strain models compared with 47% of InPec models (P = 5.0 × 10^−2). Notably, none of the EHEC strain models were able to catabolize either of these two compounds, and only 13% of Shigella models were able to use them as a sole source of carbon." See caption beneath table