Jaron C. Thompson

The paper constructs synthetic microbial communities to study interspecies interactions on nitrogen fixation and identifies a community for microbial inoculants.

The paper designs synthetic gut microbial communities using mechanistic and machine learning models to robustly inhibit C. difficile growth in vitro and in vivo.

The paper introduces a physics-constrained neural ordinary differential equation model to predict microbial community dynamics and interactions, outperforming existing mechanistic and machine learning approaches.

The paper designs human gut microbiome communities using machine learning to optimize responses to dietary fibers, identifying beneficial fiber-species combinations that enhance butyrogenic properties.

This study shows how human gut microbiota interactions influence the long-term growth and evolutionary adaptations of C. difficile, including metabolic shifts and reduced disease severity.

This study identifies gut microbiota interactions that inhibit C. difficile strains across nutrient environments, highlighting strain variability and nutrient context as key factors.

The paper develops a high-throughput optogenetic platform combined with machine learning to enable multiplexed control of light-sensitive transcription factors in yeast.

The paper develops a physically constrained recurrent neural network and Bayesian experimental design framework to optimize microbial community functions by efficiently navigating complex design spaces.

This study shows that controlling the arginine dihydrolase pathway in E. coli affects gut community assembly and butyrate production in vitro and in mice.

The paper uses LSTM to model synthetic human gut microbiome dynamics, outperforming traditional models and identifying key microbial drivers of metabolite production.

This paper develops a model-guided framework to design high-diversity synthetic human gut communities by optimizing culture conditions and leveraging ecological models.

The paper develops a machine learning tool to predict dissolved organic carbon from soil microbiome data and finds that it performs comparably to traditional methods.