Innovative new tool helps understand how gut microbes interact to influence health

The gut microbiome is made up of trillions of microbes that play a vital role in keeping us healthy.  A disturbance in the balance of these microbes can contribute to a variety of health conditions, such as inflammatory bowel disease (IBD). Now, University of California San Diego researchers have developed an innovative new tool called coralME to better understand how these microbes interact with each other and their environment to influence health. The tool rapidly creates detailed genome-scale computer models of metabolism, gene and protein expression from large amounts of data. These "ME-models," as they are called, link a microbe's genome to its phenotype, or attributes.

The models can uncover how microbes respond to certain nutrients, including which nutrients will increase certain microbes and contribute to an imbalance in the microbiome and which nutrients are most favorable to microbes that are commonly found in a healthy gut. In addition, the tool predicts what nutrients favor the formation of undesired products such as allergens or toxins.

"For example, we see from the models that a microbe needs a certain amino acid, but it cannot make this amino acid itself, so it either gets it from another microbe, from the human host, or from the diet the human is eating," said Karsten Zengler, Ph.D., professor of pediatrics at UC San Diego School of Medicine. "These next generation genome-scale models provide the mechanistic basis for understanding microbial behavior in complex environments."

Using coralME, the team generated 495 ME‑models characterizing the most common gut species, something that would have taken decades or even centuries to do by hand. 

The researchers used the models to simulate how different diets affect gut bacteria. For example, they found that low-iron or low-zinc diets allow certain harmful bacteria to survive, while diets high in certain macronutrients may promote beneficial ones, effects that traditional computational models miss.

Inputting microbial expression data from actual IBD patients into the models revealed what the microbes were doing in real time, said Zengler, who is also affiliated with the Center for Microbiome Innovation at UC San Diego and an adjunct professor of bioengineering at the Jacobs School of Engineering.

"This shows what the microbes are eating, what products they are making and how they interact with other microbes and the host," he said. "Think of the models as a road map of a city. When we integrate this map with traffic information, we get the real-time status of the map - how the traffic is flowing right now."

The team found that IBD patients experience shifts in gut chemistry, including heightened (less acidic) pH levels, but decreased production of short-chain fatty acids that normally protect the gut. They also identified specific bacteria and interactions between groups of bacteria linked to these changes.

The coralME tool provides a fast and powerful way to turn complex genetic data into concrete predictions about how gut microbes behave and how diet and disease change their activity. The insights gained from the models could lead to new ways to diagnose and treat IBD and many other diseases that affect the microbiome, with the potential to use personalized therapies to target specific microbial activities. 

"If we can accurately predict the response of the microbiome to any disease, we can understand the link between the disease and the microbiome and find a cure for the disease," said Zengler.

He says the coralME tool has applications outside of human disease as well, as it can also be used to generate models of microbial communities found, for example, in soil, other animals, or in the ocean.

The study was published on November 20, 2025, in Cell Systems.

Additional co-authors on the study include: Juan D. Tibocha-Bonilla, Rodrigo Santibáñez-Palominos, Yuhan Weng and Manish Kumar, all at UC San Diego.