Researchers have developed an AI tool that can diagnose a range of infections and health conditions in a single step by analyzing the genetic sequencing of immune cells in blood samples.
اضافة اعلان
In a study involving around 600 participants, published in the journal Science, the tool was able to identify whether the participants were healthy or suffering from conditions such as COVID-19, type 1 diabetes, HIV, lupus, as well as whether they had recently received the flu vaccine.
Although the tool is not yet ready for clinical use, with further work to refine the approach, it could one day assist doctors in dealing with "conditions that currently lack definitive tests," according to Maxim Zaslavski, a computer scientist at Stanford University in California, who participated in the study.
Sarah Tischman, a molecular biologist at the University of Cambridge in the UK, explained, "From a real-world perspective, the promise would be having a single foundational model of the immune system where you can read everything a person has been exposed to and link it to their healthcare." She added that there are still many steps to take to achieve this in the future, but this is one step forward.
Zaslavski and his colleagues developed an AI tool that combines six machine learning models to analyze the genetic sequences encoding key areas in the receptors of B and T cells and extract patterns associated with specific diseases.
The immune system maintains a vast record of current and past diseases through its two main types of cells, B cells and T cells.
B cells produce antibodies that bind to viruses and harmful molecules, while T cells trigger other responses or kill infected cells.
When a person contracts an infection or an autoimmune condition where the body mistakenly attacks its own tissues, both B and T cells proliferate and begin producing certain surface receptors.
Sequencing the genes that encode these receptors can open up a unique record of the person’s diseases and infections.
Victor Griev, a computational immunologist at the University of Oslo, states, "The immune system is a natural diagnostic, and if we only learn how it does this, we can do it too."
Zaslavski mentions that most previous efforts focused on sequencing the genes of B or T cells alone, but "combining them to get a complete picture of immune activity gives us a more comprehensive reading of what might be happening."
