Systems approaches to the infant gut microbiota From temporal dynamics to metabolic interventions

This doctoral thesis aims to investigate the infant gut microbiota across multiple scales, ranging from temporal dynamics to metabolic interventions, using systems biology approaches. First, an R package named MicrobTiSDA was developed, integrating discrete-time Lotka-Volterra models with natural spline regression to infer temporal microbial abundance dynamics and interspecies interaction networks. Second, by combining in vitro cultivation experiments with multi-strain dynamic flux balance analysis (msdFBA) based on genome-scale metabolic models (GEMs), the cross-feeding mechanisms of gut microbes driven by human milk oligosaccharides (HMOs) were elucidated in depth. The findings demonstrate that the combination of Bifidobacterium infantis and the butyrate-producing Anaerostipes caccae effectively promotes butyrate synthesis when utilizing neutral HMOs. Finally, personalized in silico models of the gut microbiota for preterm and healthy term infants were constructed using the BacArena platform. Simulations confirmed that a combined intervention strategy integrating probiotics with personalized dietary supplements exerts a potent synergistic effect, effectively restoring the production of short-chain fatty acids (SCFAs) in the preterm infant gut. This study deepens the mechanistic understanding of early-life infant gut microecological development patterns and lays a solid theoretical and computational foundation for developing personalized, targeted microecological interventions for high-risk infant populations.