How E. Coli Bacteria Could Help Predict COVID-19 Severity

A common bacterium could help identify severe COVID-19 cases without AI or genetic modification.

Researchers have developed an innovative way to predict whether a person with COVID-19 is likely to develop a mild or severe form of the disease by using one of the world’s most familiar bacteria—Escherichia coli (E. coli) (1^✔ ✔Trusted Source
Living bacterial reservoir computers for information processing and sensing

Go to source

).

The approach relies on the bacterium’s natural ability to respond to its environment and could pave the way for low-cost diagnostic tools, particularly in regions where access to advanced medical technologies is limited.

The study, led by researchers from INRAE, in collaboration with Grenoble Alpes University Hospital, Université Grenoble Alpes, and the French Alternative Energies and Atomic Energy Commission (CEA), was published in the journal Cell Systems.

Using Nature Instead of Artificial Intelligence

Unlike conventional diagnostic systems that depend on sophisticated machine learning algorithms or genetic engineering, the new method uses E. coli exactly as it exists in nature.

The researchers took advantage of the bacterium’s remarkable ability to detect and respond to changes in its surroundings. In the natural environment, E. coli constantly monitors available nutrients and other chemical signals, adjusting its metabolism and growth to survive. The scientists realized this natural behavior could also be used to process complex biological information.

Rather than programming the bacterium or modifying its genes, they allowed it to “read” the chemical composition of blood plasma collected from COVID-19 patients.

Reading COVID-19 Risk Through Bacterial Growth

In the study, E. coli bacteria were placed directly into plasma samples taken from patients infected with COVID-19.

Every plasma sample has a unique chemical signature influenced by the body’s immune response and overall health. As the bacteria encountered these differences, they altered their growth rate. Some samples caused the bacteria to grow faster, while others slowed their growth.

Researchers recorded these changes as growth curves, which reflected how the bacteria responded to the mixture of biological signals present in each patient’s plasma.

By analyzing these growth patterns, the researchers were able to distinguish patients who were likely to develop mild COVID-19 from those at greater risk of progressing to severe disease.

Remarkably, the bacteria accomplished this without any prior “training” or artificial programming, performing tasks that are typically assigned to machine-learning systems.

A Low-Cost Tool for Resource-Limited Settings

One of the biggest advantages of the new technique is its simplicity.

Many existing diagnostic and prognostic methods require specialized laboratory equipment, expensive instruments, or complex computational analysis. In contrast, the bacterial approach relies on the natural behavior of a living organism, making it potentially far less expensive and easier to deploy.

Researchers believe this could be especially valuable in low- and middle-income countries or healthcare settings where access to advanced diagnostic technologies remains limited.

An affordable test that can identify patients at higher risk of severe illness could help healthcare providers prioritize treatment, allocate hospital resources more effectively, and intervene earlier when needed.

Potential Applications Beyond COVID-19

The researchers believe the technology has applications that extend well beyond the pandemic.

They are now exploring whether the same approach can be used to analyze other clinical samples and monitor environmental samples, including urban wastewater. Because bacteria naturally respond to changes in their surroundings, they may serve as biological sensors capable of detecting complex chemical patterns in a wide range of settings.

Such applications could support disease surveillance, environmental monitoring, and public health programs without relying on costly analytical technologies.

A New Role for Living Organisms

The study highlights a growing field of biotechnology that uses living organisms as information-processing systems.

Instead of treating bacterial growth as simply a biological process, researchers demonstrated that it can function as a form of natural computation, converting the complex chemical composition of a sample into meaningful diagnostic information.

A Simpler Approach Without Genetic Modification

Importantly, the bacteria were not genetically modified, making the approach simpler and potentially easier to implement than many synthetic biology techniques.

Although additional research is needed before the technology becomes part of routine clinical practice, the findings demonstrate that naturally occurring bacteria can perform sophisticated analytical tasks previously thought to require artificial intelligence.

If validated in larger studies, the technique could offer healthcare systems an affordable and accessible way to predict COVID-19 severity while opening the door to new diagnostic applications in medicine and environmental science.

The researchers say their work represents a new way of thinking about diagnostics—one that harnesses the natural intelligence of living organisms to transform complex biological information into practical healthcare solutions.

Reference:

1. Living bacterial reservoir computers for information processing and sensing – (https://www.cell.com/cell-systems/fulltext/S2405-4712(26)00136-5)

Source-Medindia