Identifying Engineering Targets from Transcriptomic Data Using Genome-Scale Metabolic Model

Transcriptomic analysis is often employed for developing microbial cell factories to identify current bottlenecks of product biosynthesis. Transciptomic analysis in the context of metabolic network may provide further insights into the metabolism of the cell than simple differential expression analysis. In this study, we developed a new strain design algorithm named as transcriptomics-based strain optimization tool (tSOT) which identifies overexpression targets from the transcriptomic data using genome-scale metabolic model. tSOT first uses iMAT algorithm (Zur et al. Bioinformatics 2010) to classify model reactions into active and inactive reactions based on the given transcriptomic data. Subsequently, tSOT finds overexpression targets from the set of inactive reactions to increase the product yield. We applied tSOT for antibiotic overproducer design and experimentally validated the predicted targets. As the first strain design algorithm which can utilize transcriptomic data, tSOT will open the new era of strain design algorithm.