Abstract: A series of metabolic engineering modifications are needed for high efficient microbial production of bioproducts. After about 20 years' of extensive studies, the traditional metabolic engineering methods depending on intuitive analysis of local pathways are facing big challenges in discovery of new engineering strategies and new methods from different aspects are required. In this paper, the recent research progresses, on using computational analysis of large scale metabolic networks for the design of new pathways to new products or improved production yield, was reviewed; and how a microbial strain could be modified to enable and strengthen the new pathways for efficient production of biochemicals. The methods of reconstructing metabolic network model and commonly used methods for metabolic network prediction for discovering new strategies were elaborated. Metabolic pathways designed by metabolic networks were categorized into two groups, namely guiding the strain to produce a certain product from scratch or from low to high yield. In this paper, examples of combining these two types of strategies were presented. The various problems encountered in the computational design of new metabolic pathways were discussued, and feasible solutions to address those problems were proposed to enable reliable pathway design.

Key words: metabolism, biological engineering, model, optimal design

摘要： 微生物菌种需要经过一系列代谢工程改造才能实现高产特定产品，而经过二十多年的研究积累，应用传统代谢工程方法通过局部途径分析发现菌种改造策略遇到严重瓶颈，急需从新的角度进行分析以发现新的改造策略。本文结合本实验室已有工作以及文献报道的实例，阐述了应用代谢网络模型分析发现产品合成新途径进而指导代谢工程菌种改造方面的研究进展。本文总结了代谢网络模型的构建方法及应用代谢网络模型进行途径预测的常用方法，将已有的模型指导代谢途径改造的策略总结为两类，即模型指导的产品从无到有和从有到优的改造策略，并结合实例分别对这两类策略进行详细阐述。同时本文阐述了目前利用代谢网络模型进行途经预测中存在的问题，并且针对这些问题提出了一些可行的解决方案。

关键词: 代谢, 生物工程, 模型, 优化设计