Phosphoserine aminotransferase (SerC) from Escherichia coli (E. coli) MG1655 is engineered to catalyze the deamination of homoserine to 4‐hydroxy‐2‐ketobutyrate, a key reaction in producing 1,3‐propanediol (1,3‐PDO) from glucose in a novel glycerol‐independent metabolic pathway. To this end, a computation‐based rational approach is used to change the substrate specificity of SerC from l‐phosphoserine to l‐homoserine. In this approach, molecular dynamics simulations and virtual screening are combined to predict mutation sites. The enzyme activity of the best mutant, SerCR42W/R77W, is successfully improved by 4.2‐fold in comparison to the wild type when l‐homoserine is used as the substrate, while its activity toward the natural substrate l‐phosphoserine is completely deactivated. To validate the effects of the mutant on 1,3‐PDO production, the “homoserine to 1,3‐PDO” pathway is constructed in E. coli by coexpression of SerCR42W/R77W with pyruvate decarboxylase and alcohol dehydrogenase. The resulting mutant strain achieves the production of 3.03 g L−1 1,3‐PDO in fed‐batch fermentation, which is 13‐fold higher than the wild‐type strain and represents an important step forward to realize the promise of the glycerol‐independent synthetic pathway for 1,3‐PDO production from glucose.