The use of phages as antibacterials is becoming more and more common in Western countries. However, a successful phage-derived antibacterial treatment needs to account for additional features such as the loss of infective virions and the multiplication of the hosts. The parameters critical inoculation size (V F ) and failure threshold time (T F ) have been introduced to assure that the viral dose (V ϕ) and administration time (T ϕ) would lead to the extinction of the targeted bacteria. The problem with the definition of V F and T F is that they are non-linear equations with two unknowns; thus, obtaining their explicit values is cumbersome and not unique. The current study used machine learning to determine V F and T F for an effective antibacterial treatment. Within these ranges, a Pareto optimal solution of a multi-criterial optimization problem (MCOP) provided a pair of V ϕ and T ϕ to facilitate the user's work. The algorithm was tested on a series of in silico microbial consortia that described the outgrowth of a species at high cell density by another species initially present at low concentration. The results demonstrated that the MCOP-derived pairs of V ϕ and T ϕ could effectively wipe out the bacterial target within the context of the simulation. The present study also introduced the concept of mediated phage therapy, where targeting booster bacteria might decrease the virulence of a pathogen immune to phagial infection and highlighted the importance of microbial competition in attaining a successful antibacterial treatment. In summary, the present work developed a novel method for investigating phage/bacteria interactions that can help increase the effectiveness of the application of phages as antibacterials and ease the work of microbiologists.
Keywords: Pareto optimization; antibacterial treatment; machine learning; microbial ecology models; phage therapy.
Copyright © 2022 Plunder, Burkard, Lauer, Venturelli and Marongiu.