Key Questions Answered
Question: What did the ME/CFS study discover? A: Researchers found that ME/CFS alters critical links between gut bacteria, immune responses, and metabolic processes. Using this insight, they pinpointed biological markers capable of separating patients from healthy individuals with an impressive 90% diagnostic accuracy.
Question: How does the BioMapAI AI platform help? A: BioMapAI integrates thousands of data points (including microbial profiles, blood tests, immune markers, and symptoms) to identify patterns and changes specific to ME/CFS, making precision medicine more feasible.
Question: Why are these findings important for patients? A: The research not only strengthens the biological validity of ME/CFS but also offers personalized insights into the origins of symptoms. This could guide future dietary, lifestyle, and therapeutic interventions, especially for long-term COVID and other related conditions.
Summary: An advanced AI-powered investigation has uncovered that myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) profoundly alters the intricate interplay between the immune system, gut microbiota, and metabolic pathways, offering fresh insight into the biological underpinnings of the condition.
The new platform, BioMapAI, achieved 90% accuracy in identifying ME/CFS patients based on stool, blood, and symptom data, providing long-awaited confirmation for the millions of people living with this debilitating condition.
The researchers found that patients exhibited various biological characteristics, such as low levels of beneficial fatty acids, altered immune cell activity, and metabolic imbalances. These findings could guide personalized treatment and lay the scientific foundation for future treatments, especially for patients with long-term COVID-19 and similar symptoms.
Important facts:
- Advances in AI: BioMapAI distinguished between patients with ME/CFS based on immunological, microbiological, and metabolic data with 90% accuracy.
- Biological signature: Patients were found to have altered tryptophan metabolism, inflammatory responses to immune cells, and decreased butyric acid levels.
- Possibility of health-related medicine: The findings could lead to targeted interventions for ME/CFS and prolonged COVID.
Scource: Jackson Laboratory
Millions living with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS), a debilitating illness long overlooked due to the absence of reliable diagnostic tools may be on the brink of more personalized care.
New research reveals how ME/CFS disrupts the intricate interactions between the gut microbiome, immune system, and metabolism, offering fresh biological insights that could pave the way for targeted diagnostics and treatments.
The research potentially relevant to long COVID due to its similarities with ME/CFS was based on data from 249 participants analyzed using a novel artificial intelligence (AI) platform. This system identifies disease biomarkers from stool, blood, and other routine laboratory tests, offering a powerful new diagnostic approach.
“According to study author Dr. Daria Ontmaz, professor of immunology at The Jackson Laboratory, the AI achieved 90% accuracy in distinguishing individuals with chronic fatigue syndrome. She emphasized that this is a significant advance, as clinicians currently lack reliable biomarkers for diagnosing the condition”. Some doctors doubt whether it is a real disease because there are no clear biological characteristics. Sometimes they attribute it to psychological factors.
The study was spearheaded by Dr. Julia Oh, a former JAX Fellow who now serves as a microbiologist and professor at Duke University. She worked in close collaboration with ME/CFS clinicians Lucinda Bateman and Suzanne Vernon of the Bateman Horne Center, alongside Dr. Derya Unutmaz, director of the JAX ME/CFS Collaborative Research Center. The full findings are detailed in the latest issue of Nature Medicine.
Mapping the invisible
Chronic fatigue syndrome is characterized by severe symptoms that significantly affect physical and mental functioning. Common symptoms of ME/CFS include persistent fatigue, disrupted sleep, dizziness, and chronic pain. Experts frequently draw parallels between ME/CFS and long COVID, as both conditions can develop after viral infections such as Epstein–Barr virus.
In the United States, it is estimated that 836,000 to 3.3 million people are affected, yet the majority remain undiagnosed. According to the Centers for Disease Control and Prevention, the illness imposes a substantial economic burden, costing between $18 billion and $51 billion annually in healthcare expenses and lost productivity. Previous studies have observed immune changes in ME/CFS, says Unutmaz.
The team linked these associations to 12 categories of patient-reported symptoms, which were collected from hundreds of data points obtained from surveys about patients’ health and lifestyle. To carry out the study, scientists drew on a rich dataset gathered by the Bateman Horne Center in Salt Lake City, Utah—a leading institution for research into ME/CFS, long COVID, and fibromyalgia. This comprehensive resource provided the foundation for an in‑depth, multi‑year analysis.
Lead author Dr. Ruoyun Xiong developed BioMapAI, a deep neural network model designed to integrate diverse biological and clinical data. Over a four‑year span, the tool analyzed gut metagenomics, plasma metabolomics, immune cell profiles, blood test results, and detailed symptom records from 153 patients and 96 healthy individuals, enabling the identification of complex patterns linked to disease biology.
“Our data suggests that these biological changes persist over time,” Unutmaz says. “This doesn’t mean that ME/CFS is irreversible in the long term, but it may be more difficult.” Microbial imbalances were seen in patients with elevated tryptophan, benzoate, and other markers.
An Actionable Dataset
The authors note that while further validation is needed, the results represent a major step forward in understanding ME/CFS.
Oh said that animal models cannot fully capture the complex neurological, physiological, immune, and other systemic changes observed in ME/CFS. Therefore, direct studies in humans are crucial to identify modifiable factors and develop targeted therapies.
“The microbiome and metabolism are dynamic,” Oh said. “The researchers explained that this finding opens the door to practical, personalized interventions—such as dietary changes, lifestyle adjustments, or targeted treatments—that go beyond what genomic data alone can reveal. By integrating insights from the microbiome, immune system, and metabolism, they believe it’s possible to design more effective strategies for improving patient outcomes.”
BioMapAI also achieved nearly 80% accuracy on external datasets, confirming key biomarkers from the initial group. This consistency across different datasets is surprising, according to the authors. Dr. Julia Oh explained that the team’s objective is to create a detailed map showing how the immune system interacts with gut bacteria and the metabolites they produce. By linking these complex biological relationships, the researchers aim to uncover the root causes of disease and lay the groundwork for the kind of precise, personalized medicine that has long been out of reach.
The study was a collaborative effort, with contributions from Elizabeth Aiken, Ryan Caldwell, Lena Kozhaya, and Courtney Gunter of The Jackson Laboratory, as well as Suzanne D. Vernon and Lucinda Bateman of the Bateman Horne Center.
Abstract
AI-powered multi-omics modeling of myalgic encephalomyelitis/chronic fatigue syndrome
Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a chronic disease with a multifactorial etiology and heterogeneous symptoms. Therefore, diagnosis and treatment pose significant challenges. The team introduces BioMapAI, a supervised deep neural network designed to analyze complex biological data with high precision. This model was trained on a rich, four-year longitudinal multi‑omics dataset collected from 249 participants, providing a robust foundation for uncovering patterns and relationships that might otherwise remain hidden.
The dataset itself is uniquely comprehensive, combining gut metagenomics, plasma metabolomics, immune cell profiling, blood laboratory results, and detailed clinical symptom records. By integrating these diverse data streams, BioMapAI can generate deeper insights into human health, disease progression, and potential diagnostic markers, paving the way for more personalized and predictive medical approaches.
By simultaneously modeling these different data types to predict clinical severity, BioMapAI identifies biomarkers specific to the disease and symptoms and classifies ME/CFS into independent, excluded outgroups. The researchers applied an explainable artificial intelligence framework to build a detailed connectivity map linking the microbiome, immune system, and plasma metabolome. This map was generated for both healthy individuals and those diagnosed with ME/CFS, ensuring that the analysis accounted for variables such as age, sex, and other relevant clinical factors.
By integrating these diverse biological layers, the approach provides a clearer picture of how these systems interact differently in health and disease. This not only enhances understanding of ME/CFS but also offers a foundation for identifying potential biomarkers and therapeutic targets. This map reveals altered relationships between microbial metabolism (e.g., short-chain fatty acids, branched-chain amino acids, tryptophan, benzoate), plasma lipids, and bile acids, and increased inflammatory responses in mucosal inflammatory T-cell subsets (MAIT, and GN.γδA).
BioMapAI provides unprecedented systems-level insights into ME/CFS, refining existing hypotheses and raising questions about unique mechanisms, particularly how multi-omics dynamics relate to the heterogeneous symptoms of the disease.
