A systematic evaluation of Mycobacterium tuberculosis Genome-Scale Metabolic Networks

Metabolism underpins the pathogenic strategy of the causative agent of TB, Mycobacterium tuberculosis (Mtb), and therefore metabolic pathways have recently re-emerged as attractive drug targets. A powerful approach to study Mtb metabolism as a whole, rather than just individual enzymatic components, is to use a systems biology framework, such as a Genome-Scale Metabolic Network (GSMN) that allows the dynamic interactions of all the components of metabolism to be interrogated together. Several GSMNs networks have been constructed for Mtb and used to study the complex relationship between the Mtb genotype and its phenotype. However, the utility of this approach is hampered by the existence of multiple models, each with varying properties and performances. Here we systematically evaluate eight recently published metabolic models of Mtb-H37Rv to facilitate model choice. The best performing models, sMtb2018 and iEK1011, were refined and improved for use in future studies by the TB research community.Author summary: The tuberculosis bacillus, Mycobacterium tuberculosis (Mtb), is a global killer causing millions of deaths every year and is therefore a major burden to human health. Treatment of tuberculosis requires a cocktail of antibiotics for a minimum of 6 months. Treatment failure is common and is a major driver in the upward trend of antibiotic resistance, recognized by the World Health Organization as one of top ten threats to global health. A key to the success of Mtb as a human pathogen is ascribed to its extraordinary metabolic flexibility. Understanding the metabolism of Mtb is therefore an important goal of TB researchers as metabolic pathways present attractive drug targets. A powerful approach to study metabolism is through the use of genome-scale metabolic networks which enable metabolism to be studied at the whole system level rather than one enzyme at a time. Here, we comprehensively compare available genome scale metabolic networks. Our results identify the best performing networks for a variety of modelling approaches. This work allowed us to refine these models for the TB community to use in future studies to probe the metabolism of this formidable human pathogen.