ANTIBIOTIC-INDUCED PERTURBATION OF QUORUM SENSING IN THE GUT MICROBIOME: A MOLECULAR COMMUNICATION CHANNEL PERSPECTIVE

The human gut microbiota can coordinate collective microbial behaviour through interspecies quorum sensing mediated by Autoinducer-2 (AI-2). This thesis presents a computational extension of a stochastic quorum sensing model coupling prescribed bacterial population densities, extracellular AI-2 dynamics, and intracellular monitor-protein responses in the two dominant phyla, Firmicutes and Bacteroidetes. The framework introduces non-normalized, LuxS-informed production weights that represent the uneven genomic prevalence of the AI-2 synthesis pathway between the two phyla. A reference no-compensation condition is compared with three alternative Bacteroidetes compensation regimes. Two exploratory antibiotic-associated population scenarios, motivated by beta-lactam and fluoroquinolone-associated dysbiotic transitions, are examined across five transient-depression amplitudes over a 5400 s comparison interval, using ten stochastic realizations. Trajectory analysis combined with slidingwindow mutual information indicates that, in the absence of Bacteroidetes compensation, a dysbiotic transition towards Bacteroidetes can reduce community-scale AI-2 availability. In the fluoroquinolone-associated scenario, extracellular AI-2 remains below its initial eubiotic concentration for all amplitudes considered, whereas the beta-lactam-associated transition supports a late increase in AI-2 above its initial concentration. The perturbation amplitude modulates the timing of the transient molecular response and the contemporaneous AI-2–MP association, which reflects the dynamic trajectory phase and response history rather than the instantaneous extracellular concentration alone. These results provide an exploratory computational framework for characterising how composition and abundance shifts may alter quorum-sensing dynamics; they do not demonstrate clinical recovery or provide experimental validation.