Modeling regulatory networks using machine learning for systems metabolic engineering

Abstract

Systems metabolic engineering attempts to engineer a production host's biological network to overproduce valuable chemicals and materials in a sustainable manner. In contrast to genome-scale metabolic models that are well established, regulatory network models have not been sufficiently considered in systems metabolic engineering despite their importance and recent notable advances. In this paper, recent studies on inferring and characterizing regulatory networks at both transcriptional and translational levels are reviewed. The recent studies discussed herein suggest that their corresponding computational methods and models can be effectively applied to optimize a production host's regulatory networks for the enhanced biological production. For the successful application of regulatory network models, datasets on biological sequence-phenotype relationship need to be more generated.