Data_Sheet_1_Determination of phage load and administration time in simulated occurrences of antibacterial treatments.docx

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 (VF) and failure threshold time (TF) have been introduced to assure that the viral dose (Vϕ) and administration time (Tϕ) would lead to the extinction of the targeted bacteria. The problem with the definition of VF and TF 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 VF and TF for an effective antibacterial treatment. Within these ranges, a Pareto optimal solution of a multi-criterial optimization problem (MCOP) provided a pair of Vϕ and Tϕ 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 Vϕ and Tϕ 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.

噬菌体作为抗菌剂在西方国家的应用日趋普及。然而，成功的噬菌体源性抗菌治疗需综合考量诸多额外因素，例如感染性病毒粒子的损耗与宿主的增殖行为。学界已引入临界接种量（critical inoculation size, VF）与失效阈值时间（failure threshold time, TF）两个核心参数，旨在确保病毒剂量（viral dose, Vϕ）与给药时间（administration time, Tϕ）可实现目标细菌的彻底根除。但VF与TF的定义存在固有局限：二者对应的控制方程为二元非线性方程组，因此求取其显式解不仅过程繁琐，且解不唯一。本研究采用机器学习方法，确定了可实现有效抗菌治疗的VF与TF最优取值范围。在此参数范围内，多准则优化问题（multi-criterial optimization problem, MCOP）的帕累托最优解可输出一组Vϕ与Tϕ参数组合，以简化科研人员的工作流程。该算法已在一系列硅基模拟微生物群落（in silico microbial consortia）中完成测试，此类群落模拟了初始浓度较低的物种在高细胞密度环境下，取代原有物种并实现自身增殖的过程。实验结果显示，基于MCOP得到的Vϕ与Tϕ参数组合可在模拟环境中有效根除目标细菌。本研究同时提出了介导噬菌体疗法的概念：通过靶向助推细菌（booster bacteria），可降低对噬菌体感染免疫的病原体的毒力，并强调了微生物竞争在实现成功抗菌治疗中的关键作用。综上，本研究开发了一种探究噬菌体-细菌相互作用的全新方法，可助力提升噬菌体作为抗菌剂的应用效能，并简化微生物学家的科研工作。