Artificial Intelligence discovers new antibiotics against superbugs

Artificial Intelligence is becoming a revolutionary tool in antibiotic development. While neural networks were used only to search for promising molecules a few years ago, they now help understand the mechanisms of action for new drugs. This could significantly accelerate the creation of new treatments at a time when drug-resistant bacteria are spreading. This is reported by Ixbt.com reports .

A team led by Jonathan Stokes from McMaster University tested nearly 10,000 compounds against Escherichia coli and identified a new antibiotic called enterololin. Researchers used the DiffDock system, created by MIT expert Regina Barzilay, to prove that this substance affects only pathogens without harming beneficial microflora. The algorithm predicted which proteins the molecule interacts with, allowing for a significant reduction in laboratory experiments.

The search for new antibiotics is currently a highly urgent issue. It is estimated that by 2050, drug-resistant infections could cause 39 million deaths worldwide. Since this is an expensive and low-profit field for pharmaceutical companies, Artificial Intelligence serves as a key solution to reduce costs and accelerate the process.

Previously, the MIT team discovered the antibiotic halicin using the Chemprop system. Meanwhile, the APEX system, led by Cesar de la Fuente from the University of Pennsylvania, analyzed over 10 million protein molecules and found more than 37,000 potential antibiotics. Interestingly, some of them were found in proteins from ancient plants and extinct animals, attacking modern bacteria in completely new ways.