It’s a first. A new class of antibiotics has just been discovered using artificial intelligence software. It would be effective against drug-resistant Staphylococcus aureus bacteria (MRSA).
This is important news for the medical world and a first in sixty years. A new class of antibiotics, effective against drug-resistant Staphylococcus aureus (MRSA) bacteria, has just been discovered by artificial intelligence. The work was carried out by American researchers and published in the scientific journal Nature.
A discovery based on artificial intelligence
Using artificial intelligence algorithms, the team of scientists from the Broad Institute of MIT and Harvard tested the effects of more than 39,000 compounds on Staphylococcus aureus and on three types of human cells (liver, skeletal muscle and lungs) with the aim of evaluating both their antibiotic properties and their toxicity on human cells.
A test on more than 39,000 compounds
The results of these tests as well as the chemical structures of the 39,000 compounds made it possible to provide data to artificial intelligence. After this first step, the AI was able to study 12 million substances already commercially available. This is how scientists were able to identify compounds from five different families of substances, each with their respective chemical structures.
After other laboratory tests and a selection of important molecules, two candidates from the same class were selected and tested on mice. Tests with encouraging results, because the substances managed to divide the MRSA population by 10.
An important step forward against antibiotic resistance
An important advance, because due to the significant exposure of bacteria to antibiotics, they have developed antibiotic resistance. A serious phenomenon, responsible for 1.2 million deaths in 2019. “Our model not only tells us which substances have an antibiotic effect. It also shows us why their chemical structure gives them this property.” explains Felix Wong, postdoctoral researcher at the Broad Institute of MIT and Harvard, and author of the publication in Nature.
Scientists are counting on this new model not only to discover new drugs but also to identify new potential targets for these treatments.