Duke Engineers Use Robots and AI to Improve Gut Health

Biomedical engineers at Duke University have demonstrated a method for systematically developing novel, complex combinations of probiotics and prebiotics to more effectively maintain gut health and treat gastrointestinal diseases, according to findings published online July 27 in the journal Nature Chemical Biology. By combining robotic automation with machine learning, the approach addresses longstanding consistency challenges in the estimated $130 billion global gut health industry by identifying specific bacterial combinations and dietary fibers that reliably promote human gut function despite high individual variability in diet, genetics, and microbiome composition.

Overcoming Microbiome Variability with Bayesian Optimization

Designing effective gut health treatments has historically been hindered by the vast complexity of the human gastrointestinal tract. According to Duke University researchers, the human gut hosts thousands of microbial species whose cells roughly equal the number of human cells, weighing up to nearly half a pound. This composition varies greatly by region, person, diet, and lifestyle, making standard probiotic interventions unpredictable. “Probiotics alone might not be able colonize a person’s gut long enough to do anything helpful because there’s so much living there already,” said Ophelia Venturelli, associate professor of biomedical engineering at Duke University. To overcome this, the research team deployed an approach known as Bayesian optimization, or active learning, closing the loop between computer models and automated benchtop experiments to navigate trillions of potential microbial and dietary combinations.

Robotically Automating High-Throughput Gut Experiments

To build predictive accuracy, Venturelli’s team used autonomous robotic systems to conduct high-throughput experiments evaluating interactions between 15 different species of gut microbes and six types of dietary fibers known to affect the production of butyrate, a short-chain fatty acid critical to gut health. By processing up to 390 conditions simultaneously across five batches of experiments, the automated platform rapidly filled knowledge gaps in the computer model. “We found a lot of unexpected effects from interactions between the microbes and the fibers,” Venturelli noted, adding that these complex dynamics could not have been predicted without empirical testing.

Identifying Reliable Synergies for Gut Health

The high-throughput screening revealed specific functional combinations that successfully bypassed individual microbiome variability. According to the study results, a combination of the dietary fiber inulin alongside two inulin-hungry microbial species—Bacteroides uniformis and Anaerostipes caccae—interacted with a third bacterial species, Prevotella copri, to reliably produce the desired butyrate. Crucially, this specific metabolic output held strong regardless of what other species and background factors were introduced to the culture.

Building on these benchtop discoveries, the research team is actively investigating translational applications. According to the study authors, researchers are currently testing whether this validated bacterial and fiber combination can effectively treat inflammatory bowel disease using a mouse model, with promising early results. Beyond this specific formulation, the computational and robotic framework is designed to discover tailored solutions for a wide range of gastrointestinal disorders. “There are companies combining bacteria and dietary fibers in their gut health products already, but they are not yet trying to identify and match the right synergies to the right problem,” Venturelli stated. Funding for this research was provided by the National Institutes of Health under grants R35GM124774, R01EB030340, and R01DK13346, as well as the Army Research Office under grant W911NF-19-1-0269.

Frequently Asked Questions

What is the main challenge of traditional probiotics?

Traditional probiotics often struggle to colonize the human gut long enough to provide a benefit because the gastrointestinal tract already hosts a dense, highly variable ecosystem of native microorganisms, according to Duke University researchers.

How does Bayesian optimization improve microbiome research?

Bayesian optimization uses machine learning to design targeted experiments that fill gaps in computer models, allowing researchers to efficiently navigate trillions of potential combinations between microbes and dietary sources using automated robotic systems.

What specific combination was found to reliably produce butyrate?

According to the study published in Nature Chemical Biology, the dietary fiber inulin combined with Bacteroides uniformis, Anaerostipes caccae, and Prevotella copri successfully produced butyrate consistently regardless of other introduced species.