Curating A Genome-Scale Metabolic Model of Mycoplasma Gallisepticum Using An Evolutionary Algorithm

Mycoplasma gallisepticum is a bacterial pathogen that causes chronic respiratory disease in chickens and infectious sinusitis in turkeys. The disease detrimentally impacts the poultry industry through the deaths of chicks and poult, lowered egg production, and high drug costs.  Economically, the disease results in annual costs ranging from hundreds of millions to billions of dollars in damage.  Through better understanding of this pathogen, it may be possible to identify new treatment strategies, develop better drugs, or create new vaccines.

As a first step in developing a strategy to mitigate the damage caused by M. gallisepticum, a genome-scale metabolic model was developed.  Using a number of different sources, it was possible to combine information based on the bacteria’s genome annotation, bioenergetic requirements, as well as its associated reaction stoichiometry to create a preliminary model.  However, even with this information, it was not possible to generate a feasible solution for the model.  It was hypothesized that metabolic constraints were either missing or incomplete.  To determine which metabolic constraints were incomplete, a genetic algorithm was implemented to evolve a metabolic model that resulted in a feasible solution.  After carrying out 34 runs of 2,000 generations each, it was determined that by dropping 5 of the 442 metabolic constraints, it was possible to create a viable genome-scale metabolic.  These results suggest that the metabolites that had to be dropped are likely involved in other unaccounted for reactions.