OptDesign: Identifying Optimum Design Strategies in Strain Engineering for Biochemical Production

Computational
tools have been widely adopted for strain optimization
in metabolic engineering, contributing to numerous success stories
of producing industrially relevant biochemicals. However, most of
these tools focus on single metabolic intervention strategies (either
gene/reaction knockout or amplification alone) and rely on hypothetical
optimality principles (e.g., maximization of growth) and precise gene
expression (e.g., fold changes) for phenotype prediction. This paper
introduces OptDesign, a new two-step strain design strategy. In the
first step, OptDesign selects regulation candidates that have a noticeable
flux difference between the wild type and production strains. In the
second step, it computes optimal design strategies with limited manipulations
(combining regulation and knockout), leading to high biochemical production.
The usefulness and capabilities of OptDesign are demonstrated for
the production of three biochemicals in Escherichia
coli using the latest genome-scale metabolic model
iML1515, showing highly consistent results with previous studies while
suggesting new manipulations to boost strain performance. The source
code is available at https://github.com/chang88ye/OptDesign.