A.I. ‘digital twin’ microbiome model may predict infant neurodevelopment

Using data from fecal samples collected from preterm infants in the neonatal intensive care unit (NICU), researchers from the University of Chicago report that their model, called Q-net, can predict which babies were at risk for cognitive deficits with 76% accuracy.

The gut microbiome is known to have a profound impact on the health and development of infants, but understanding how gut bacteria interact, and how these interactions may lead to gastrointestinal diseases and neurodevelopmental deficits, is difficult and time consuming through traditional laboratory experiments.

“You can only get so far by looking at snapshots of the microbiome and seeing the different levels of how many bacteria are there, because in a preterm infant, the microbiome is constantly changing and maturing,” said the study’s senior author Ishanu Chattopadhyay, PhD, assistant professor of medicine, in a press release. “So, we developed a new approach using generative AI to build a digital twin of the system that models the interactions of the bacteria as they change.”

The ‘digital twin’ concept is a potentially transformative technology. Writing in Science Advances​, the authors explained that a ‘digital twin’ is a digital representation of a complex system, “enabling the simulation of perturbations; study of trajectories, aberrations and failures; and the execution of high-fidelity simulation experiments that would otherwise be unattainable in the real world.

“Rather than answering a single question, a digital twin aims to mirror the entire system, distinguishing it from typically more limited standard machine learning models,” they explained.

Q-net​

Compared to typical wet lab experiments, where the investigation of two-way interactions of a typical bacterial colony with 1,000 species would take more than 1,000 years, the UChicago’s Q0net model can model the complex interactions in a fraction of this time.

Dr. Chattopadhyay and his colleagues trained the model using data obtained from 58 infant fecal samples at UChicago’s Comer Children’s Hospital. Next, they validated its predictions about how the microbiome would develop using data from 30 preterm infants at Beth Israel Deaconess Medical Center in Boston.

The results indicated a 76% accuracy in predicting which babies were at risk for cognitive deficits, as measured by head circumference growth.

The model also indicated that interventions like restoring the abundance of a particular bacterial species could reduce the developmental risk of about 45% of the babies. The authors caution, however, that the model also showed that incorrect interventions can make the risk worse.

“You can’t just give probiotics and hope that the developmental risk is going to go down,” said Dr. Chattopadhyay. “What you are supplanting is important, and for many subjects, you also have to time it precisely.”

Since Q-net can identify potentially interesting combinations of bacteria, it can vastly narrow the search for potential intervention targets, said the researchers.

Dr. Chattopadhyay’s research partners, like Erika Claud, MD, professor of pediatrics and director of the Center for the Science of Early Trajectories at UChicago, are reported to be working with bioreactors that simulate the live gut microbiome environment where they can test out potential interventions and see what happens.

The study was supported financially by the U.S. National Institutes of Health.

Source: Science Advances​
2024, Volume 10, Number 15, doi: 10.1126/sciadv.adj0400​
“A digital twin of the infant microbiome to predict neurodevelopmental deficits”
Authors: N. Sizemore et al.