Inferring optimal minimal medium on genome-scale metabolic models using evolutionary algorithms

Abstract

Genome-scale metabolic models (GSMMs) are a valuable tool for the study of metabolic systems biology through biomedical to industrial research and are becoming available for an increasing number of single organisms and more recently also for microbial communities. One of the most promising features for the use of GSMMs is the rational design of microorganisms in isolation or in communities that could turn them capable of producing desired compounds in industrially relevant amount. The metabolic engineering or design problem can be simply formulated as the maximization of the production of a target compound by manipulating either environmental conditions, performing genetic manipulations or even, in the case of a microbial community, manipulate microbial composition in terms of species. In this work, it has been implemented and validated an optimization framework that allows to find an optimal minimal medium composition for a given objective function, such as maximizing growth, or the production of a given target compound. This framework was fully implemented in Python language and the workflow of the optimization process uses Evolutionary Algorithms (EA). The code, installation files and documentation are available at the GitHub repository (https://github.com/BioSystemsUM/optimModels). For the validation of this framework it was used published GSMMs of single prokaryotic organisms and natural and synthetic microbial communities. All results were compared and validated with experimental data in literature. Overall, the results obtained for minimal medium composition using the developed tool showed biological significance, correctly predicting the minimal medium in aerobic/anerobic and light/dark conditions, as required by the specific organisms involved.