StrainOptimizer empowers rational cell factory design through multi-scale metabolic models with expression and proteome constraints

Abstract

The rational design of microbial cell factories for high bioproduction remains a key challenge in metabolic engineering. While advanced modelling frameworks incorporating protein resource allocation, such as enzyme-constrained models (ecGEMs) and Expression and Thermodynamic Flux (ETFL), provide superior predictive power, their application is limited by a lack of user-friendly computational tools. Here, we present strainOptimizer, a comprehensive computational platform for rational strain design that systematically evaluates key resource allocation principles: the coupling of gene expression with metabolism, subcellular compartmentalization, and enzyme capacity limitations. Our benchmark analyses demonstrate that each principle offers distinct advantages: models coupling metabolism and expression (like ETFL) enable the identification of non-metabolic targets, organelle-level proteomic constraints improve precision for high-protein-cost products, and protein-usage-based objectives consistently outperformed traditional flux-based approaches. To demonstrate its practical utility, we applied strainOptimizer to an engineered sclareol-overproducing Saccharomyces cerevisiae strain. The platform identified novel targets, and experimental validation confirmed a 67% success rate, increasing the final sclareol titer by 14-26% and productivity by up to 45%. StrainOptimizer bridges the gap between resource allocation theory and applied engineering, providing a powerful, validated tool to accelerate the development of high-performance cell factories.