Developing AI agents to investigate Inflammatory Bowel Disease

Inflammatory Bowel Diseases (IBD), including Crohn’s disease and ulcerative colitis, are driven in part by disruptions in the gut microbiome. Thousands of microbiome and metabolite studies have been published, but it remains difficult to convert this information into reliable, actionable insights that can guide new therapies. A key barrier is human bandwidth: integrating many complex datasets, controlling for confounders and connecting findings to biological mechanisms is slow and error-prone. We will build AI “co-pilots” that help scientists do this work rigorously and transparently. These AI agents will retrieve relevant evidence, run microbiome-specific analyses across multiple IBD studies and keep a transparent record linking each conclusion to the supporting data and sources. This approach hastens discovery while enabling human researchers to dive into, or direct, more detailed analyses if desired. The system will generate ranked, testable hypotheses about microbial species, genes and metabolites that contribute to inflammation or predict disease flares or treatment response. We will then validate top hypotheses using controlled laboratory experiments and gnotobiotic mouse models, and use the results to iteratively improve the AI. The outcome will be a practical discovery engine that accelerates microbiome-based discovery and therapeutic development for IBD.