Researchers claim to have discovered a novel form of antibiotic that is effective against a particularly dangerous drug-resistant pathogen using artificial intelligence.
The antibiotic suppressed the development of the bacteria when it was tested on the outermost layer of mice that had been artificially contaminated with the superbug, indicating that the technique could be used to develop antibiotics specifically designed to combat additional drug-resistant infections.
Additionally, the AI-identified drug functioned to inhibit the problematic virus exclusively. It was an unusual, narrowly targeted agent because it didn’t appear to harm the numerous other species of helpful bacteria in the gastrointestinal tract or skin. According to the researchers, more effective medicines could stop germs from developing resistance in their initial place.
As per the CNN report, The Acinetobacter baumanii bacteria was the main subject of the study. It lingers on surfaces like knobs on doors and countertops in hospitals and other healthcare facilities. It can exploit the most incredible weaponry of the other creatures it interacts with—genes that enable them to fend off drugs used to treat them—because it can snag chunks of DNA from them.
According to one of the researchers and assistant professor of biochemistry and biological sciences at McMaster University in Hamilton, Ontario, Jon Stokes, “It’s what we call in the laboratory a professional pathogen.”
Stokes’ lab collaborated on the latest study alongside MIT and Harvard’s Broad Institute academics. First, they grew Acinetobacter baumanii on lab plates using a method called high-throughput drug screening. Then they spent weeks subjecting the colonies to more than 7,500 agents, including medicines and drug-active ingredients. They discovered 480 substances that prevented the bacteria’s development. They entered that data into a computer and then utilized it for training an algorithm for artificial intelligence.
As soon as our model was trained, we could expose it to fresh images of compounds it had never seen before, right? And it would tell us if those compounds were antimicrobial or not based on what it had learned throughout training, according to Stokes.
Then, Stokes said, they asked the model to filter more than 6,000 compounds, which the AI achieved in a few hours. According to him, it operates entirely novelly by delaying the bacteria’s components from leaving the cell inside and reaching the cell’s surface.
According to Stokes, the majority of antibiotics are broad-spectrum drugs that fight a variety of bacterial species. Numerous varieties of bacteria are subjected to intense selection pressure by broad-spectrum antibiotics, which drives many of them to rapidly evolve and share genes that enable them to withstand the medication and thrive.
“With this molecule, it doesn’t impose that universal, selective pressure, so it’s not going to spread resistance quite as quickly,” he said. “It only works extremely potently against Acinetobacter.”