Title: Stochastic processes for dose response of antibiotic resistant bacteria
Abstract: Antibiotic resistant bacteria (ARB) are a growing problem due to rapid evolution of resistance to existing drugs and concomitant void in drug discovery. To estimate the size of their threat and manage the risk, it is useful to understand the probability of a person falling ill from a given bacterial dose. Studies producing such data involve inoculating volunteers with ARB and raise ethical concerns. Here a stochastic modeling approach is discussed that leverages data from in-vitro experiments to predict the relationship between infection probability and number of ARB. Specifically, an example of E. coli and its Gentamicin resistant strain are shown. Preliminary results for S. aureus will also be presented. Techniques used include stochastic processes, Bayesian optimization with STAN, optimization and Gillespie simulation with Python.