A significant goal in post-genome era is to relate the annotated genome-scale functions to the defined metabolism of a cell. Working with the complex and ill-defined nature of the cellular metabolism in even the simplest industrial microorganism, it is impossible to make the technologies of metabolic and cellular engineering more predictable and exact. In order to facilitate to make a better model organism and a more useful technological tool for metabolic engineering. We developed techniques for stepwise genomic engineering and applied them to deleting the most complex pathways of Escherichia. coli, amino acid biosynthesis, identified as dispensable by comparative genomics. Our method leaves no markers behind and can be applied sequentially. The pathways for 20 amino acids biosynthesis were successfully deleted, resulting in a reduced metabolic network. We found that the experimental data, in which multiple amino acid auxotroph in stationary phase under a range of nutrient limited condition outgrow wild type, were consented with the stated propose, namely that the pathway of amino acid biosynthesis is load to cell, and is dispensable. This experiential data thus demonstrates that the reduced metabolic set, which will provide a more defined and efficient backbone for metabolic engineering.