Using gut microbiota as a health compass

Machine learning model to predict potential NAFLD. Credit: Howell Leung/ Leibniz-HKI

The human microbiome can provide information about the risk of non-alcoholic fatty liver disease. This was discovered by an international team led by the Leibniz Institute for Natural Products Research and Infection Biology-Hans Knöll Institute. The researchers developed a model capable of predicting the possible course of the disease based on the microbial composition of the intestine. The study is published in Science Translational Medicine.

Up to 25% of the world’s population is affected by non-alcoholic fatty liver disease (NAFLD), in which an increased amount of fat cells form in the liver. It is the most common chronic liver disease in industrialized countries of the world and, unlike alcoholic fatty liver disease, it is not caused by heavy alcohol consumption. In some people, undetected NAFLD can lead to liver scarring, liver cancer, or liver failure.

In a long-term study, an international research team led by Gianni Panagiotou, research group leader for systems biology and bioinformatics at Leibniz HKI, analyzed stool and blood samples from 1,200 people who were initially free. from NAFLD. “Microorganisms in the human gut have already been shown to contribute to the development of NAFLD. We wanted to know if the microbiome of a healthy person could predict whether or not they would develop NAFLD in the future” , explains Panagiotou.

When the subjects were re-examined four years later, it was revealed that 90 of them had since developed NAFLD. Samples from affected individuals were compared to a control group of 90 individuals who did not have NAFLD at baseline or at the follow-up visit. “Using different methods, we were able to find very subtle differences in the samples we took four years before,” says first author Howell Leung of Panagiotou’s group at Leibniz HKI. “With these data, we were able to develop a model capable of predicting who will develop NAFLD in the future based on the microbiome with 80% certainty.”

Currently, there are clinical models that use biochemical parameters in the blood to make a prediction with 60% accuracy. “The model we developed combines easily measurable information from blood with data from the microbiome and can thus significantly increase reliability,” says Panagiotou.

Disease prediction using machine learning

The research team developed a so-called machine learning model, a computer model that is trained to recognize certain patterns in a data set. The model can then use these models to analyze new data sets and, in this case, predict possible non-alcoholic fatty liver disease. “The whole process of developing our model took more than three years due to the complexity of the data. However, we were ultimately successful and were able to create a useful tool for predicting NAFLD,” says Panagiotou.

Advanced non-alcoholic fatty liver disease is irreversible and, in the worst case, can even lead to liver cancer. People who already suffer from a precursor or who are particularly at risk must therefore be identified early in order to be able to counteract the disease. “NAFLD is a silent disease. This means that in most cases it is asymptomatic and is usually only detected by chance”, explains Gianni Panagiotou. The number of Germans suffering from NAFLD is estimated at around 12 million. People with pre-existing conditions such as type 2 diabetes, obesity, high blood pressure or dyslipidemia are particularly affected by fatty liver disease.

Possible applications and next steps

Thanks to their machine learning model, the researchers have already been able to compare and thus validate their results with patient data from the United States and Europe. In the next step, Panagiotou plans to conduct the study globally and use artificial intelligence to incorporate even larger datasets into the study.

“I see microbiome-based diagnostics as something that will reach clinical practice and have great potential in the next ten years,” says Panagiotou. Early treatment of risk factors for non-alcoholic fatty liver disease, such as type 2 diabetes, hypertension and obesity, could halt the development of the disease. Therefore, an early prognosis is the only way to prevent the disease.

Scientists identify how non-alcoholic fatty liver disease increases the risk of vascular disease

More information:
Howell Leung et al, Risk assessment with gut microbiome and metabolite markers in the development of NAFLD, Science Translational Medicine (2022). DOI: 10.1126/scitranslmed.abk0855

Provided by Leibniz-HKI

Quote: Using the Gut Microbiome as a Health Compass (June 9, 2022) Retrieved June 9, 2022 from

This document is subject to copyright. Except for fair use for purposes of private study or research, no part may be reproduced without written permission. The content is provided for information only.

Leave a Comment