0
0
March 2, 2025 – Researchers at Stanford Medicine have developed a groundbreaking machine-learning technique, Mal-ID, that can decipher the immune system’s “fingerprints” to diagnose a wide range of diseases, including autoimmune conditions like lupus and type 1 diabetes, as well as infections like COVID-19.
The study, published in Science, demonstrates that by analyzing the sequences and structures of B and T cell receptors, Mal-ID can accurately identify diseases based solely on a person’s immune profile. This approach leverages the immune system’s “biological Rolodex,” which contains a lifetime’s worth of information about encountered threats.
“The diagnostic toolkits that we use today don’t make much use of the immune system’s internal record of the diseases it has encountered,” said postdoctoral scholar Maxim Zaslavsky, Ph.D., a lead author of the study. “But our immune system is constantly surveilling our bodies with B and T cells, which act like molecular threat sensors.”
The researchers used a machine-learning algorithm, similar to those that power large language models like ChatGPT, to analyze over 16 million B cell receptor sequences and 25 million T cell receptor sequences from nearly 600 individuals. This included healthy controls and patients with various conditions, such as COVID-19, HIV, lupus, and type 1 diabetes.
Key findings of the study include:
- Mal-ID accurately identified disease states based on B and T cell receptor sequences.
- T cell receptor sequences were most informative for autoimmune diseases like lupus and type 1 diabetes.
- B cell receptor sequences were most informative for viral infections and vaccine responses.
- Combining T and B cell data improved the algorithm’s accuracy across all conditions.
The researchers believe Mal-ID has the potential to:
- Improve the diagnosis of complex and difficult-to-diagnose diseases.
- Track responses to cancer immunotherapies.
- Identify disease subcategories for personalized treatment.
- Uncover new therapeutic targets.
“Mal-ID could help us identify subcategories of particular conditions that could give us clues to what sort of treatment would be most helpful for someone’s disease state,” said Scott Boyd, MD, Ph.D., a senior author of the study.
The researchers envision Mal-ID being adapted to identify immunological signatures for numerous diseases and conditions, with a particular focus on autoimmune diseases.
Disclaimer: This news article is based on information available at the time of publication and reflects the findings of a specific study. Machine learning and immunological research are ongoing, and further studies may provide additional insights or modify current understandings. This article is for informational purposes only and does not constitute medical advice. Individuals with health concerns should consult with a qualified healthcare professional1 for diagnosis and treatment. The information regarding Mal-ID is based on information provided by Stanford Medicine.
