Scientists discover the first new antibiotics in over 60 years using AI

Technology News, World :- Artificial intelligence (AI) is reshaping the landscape of medicine, and its latest triumph comes in the form of a groundbreaking discovery—a novel class of antibiotics designed to combat drug-resistant Staphylococcus aureus (MRSA) bacteria. This breakthrough, made possible by transparent deep learning models, marks the first instance of new antibiotics emerging in six decades.

Led by James Collins, a professor of Medical Engineering and Science at the Massachusetts Institute of Technology (MIT), the research team harnessed the power of deep learning to unravel the mysteries of antibiotic development. The innovative approach allowed scientists to discern how the AI models learned and predicted the efficacy of certain molecules as potential antibiotics.

"This insight provided us with a time-efficient, resource-efficient, and mechanistically insightful framework, offering unprecedented understanding from a chemical-structure standpoint," explained Collins.

Published in Nature, the study employed a team of 21 researchers aiming to demystify the traditional "black box" of deep learning models. The team focused on utilizing an extensively expanded deep learning model, trained with a vast dataset comprising approximately 39,000 compounds evaluated for their antibiotic activity against MRSA.

To refine the selection of potential drugs, the researchers introduced three additional deep-learning models, assessing the toxicity of compounds on different types of human cells. By integrating toxicity predictions with antimicrobial activity, the team identified compounds capable of effectively combating microbes with minimal harm to human cells.

The screening process involved evaluating around 12 million commercially available compounds, resulting in the identification of compounds from five distinct classes with predicted activity against MRSA. Subsequent laboratory tests confirmed the effectiveness of these compounds, leading to the discovery of two promising antibiotic candidates from the same class.

In experiments involving mouse models for MRSA skin and systemic infections, each of these compounds demonstrated a remarkable reduction in the MRSA population. This groundbreaking approach not only opens new avenues in antibiotic discovery but also provides a blueprint for future advancements in the battle against antibiotic resistance.

The study success underscores the transformative potential of AI in medicine, paving the way for a new era in the fight against infectious diseases.