Researchers have created an artificial intelligence (AI) tool capable of diagnosing various infections and health conditions in a single analysis by examining immune-cell gene sequences from blood samples. In a recent study involving nearly 600 participants published in *Science* on February 20, the tool was able to determine whether individuals were healthy or affected by COVID-19, type 1 diabetes, HIV, or the autoimmune disease lupus, as well as identify those who had recently been vaccinated against the flu. “This is a one-shot sequencing method that captures all the exposures of your immune system, ” states Sarah Teichmann, a molecular biologist at the University of Cambridge, UK. Though the tool is not yet suitable for clinical applications, further refinement could enable it to assist clinicians in diagnosing “conditions that currently lack definitive tests, ” notes study co-author Maxim Zaslavsky, a computer scientist from Stanford University in California. “From a practical perspective, the goal would be to have a unified model for the immune system that can provide insights into all of a person's exposures and connect that information to their healthcare, ” explains Teichmann. “While there are many steps needed to achieve this in the future, we have made progress with this initial step. ” Natural Diagnostic Capability The immune system keeps a comprehensive record of past and present illnesses via its two primary cell types: B cells and T cells. B cells create antibodies to target viruses and harmful substances, while T cells trigger additional immune responses or destroy infected cells. When an individual suffers from an infection or an autoimmune disorder in which the body erroneously attacks its tissues, their B cells and T cells proliferate and begin producing specific surface receptors.

Analyzing the genes responsible for these receptors can reveal a person's unique history of diseases and infections. “The immune system is an inherent diagnostic tool, and if we understand how it functions, we could replicate that process, ” says Victor Greiff, a computational immunologist at the University of Oslo. Current diagnostic methods “utilize the immune system's disease exposure record to a limited extent, ” mentions Zaslavsky, but prior approaches have mainly concentrated on sequences from either B or T cells. “Integrating data from both provides a comprehensive view of immune activity and a better understanding of potential health issues. ” Rapid sequencing techniques accelerate genomic diagnosis Zaslavsky and his team developed an AI tool that combines six machine-learning models to analyze gene sequences related to critical regions in B-cell and T-cell receptors, identifying patterns linked to specific diseases.