In 60 years first new class of antibiotics identified by AI to treat multidrug-resistant infections


Researchers from prestigious institutions MIT, Harvard, and the Broad Institute have achieved a significant breakthrough in antibiotic discovery using artificial intelligence. Their explainable AI approach identified the first new class of antibiotics in over 60 years to treat drug-resistant staph infections.

The researchers developed an innovative AI technique called graph neural networks to screen a massive database of over 12 million chemical compounds. By analyzing the molecular structures, the AI models predicted which compounds could effectively inhibit the growth of methicillin-resistant Staphylococcus aureus (MRSA).

MRSA is a dangerous staph infection that has become resistant to common antibiotics. The AI models were also trained to screen for toxicity to human cells. From the vast chemical library, the AI identified promising antibiotic candidates.

After testing in the lab, the researchers found one novel compound that showed potent ability to treat MRSA, both through topical skin application and systemic treatment in mouse models. This demonstrates the compound’s viability as a new antibiotic medicine.

This pioneering AI breakthrough provides hope in the race against rising antibiotic resistance worldwide. Drug-resistant superbugs are a major global health threat, with over 120,000 deaths from MRSA alone in 2019. Experts warn that without new medicines, antibiotic resistant infections could cause 10 million deaths per year by 2050.

The researchers believe their explainable AI technique can accelerate the discovery of new classes of life-saving antibiotics. This approach analyzing molecular structures can unlock new antimicrobial medicines. Their study serves as an exciting model for AI-enabled drug development.

This first new antibiotic in over half a century demonstrates the immense potential of artificial intelligence in medicine and therapeutic discovery. The researchers’ explainable AI methodology could help address the antibiotic resistance crisis looming over modern healthcare. Their discovery offers new hope in fighting drug-resistant superbugs and saving lives from dangerous infections using the transformative power of AI. Continue reading in Springer Nature (link).