Revolutionary AI unveils new antibiotics: a historic breakthrough after 60 years of waiting!

Show summary Hide summary

New advancement thanks to AI: An antibiotic revolution

Artificial intelligence (AI) is currently revolutionizing various fields, notably health. The proof is with the appearance of a new class of antibiotics, brought to light by this technology after more than half a century without major innovation in this field.

A giant step against resistant bacteria

Advanced artificial intelligence models have paved the way for the detection of an innovative agent capable of combating multi-drug-resistant Staphylococcus aureus (MRSA), a bacterial strain responsible for thousands of deaths annually worldwide . This breakthrough could be decisive in the fight against growing resistance to existing antibiotics.

Prediction of high-performance molecules

James Collins of MIT, co-author of the study, explains that the objective was to probe the relevance of AI models for predicting the potential effectiveness of certain molecules as antibiotics. This experimental framework offers a more economical and efficient method that examines the chemical structure of compounds in a previously unprecedented way.

Promising results published in the journal “Nature”

The conclusions of this research are now accessible in the scientific journal “Nature”. The article summarizes the work of 21 researchers who contributed to this study.

To read The shocking revelation of Google’s AI: images of the American mission to the Moon called into question, embarrassing Putin!

Towards a better understanding of AI models

MIT scientists sought to “open the black box” of artificial neural networks, used by AI to learn and generate innovative ideas. The approach adopted was based on a deep learning model capable of predicting antibacterial effectiveness and non-toxicity in humans.

The compound selection method

It was through an initial filtering of some 39,000 substances for their action against MRSA, followed by a detailed analysis of their chemical structures, that the MIT team was able to feed its AI model. Felix Wong, researcher at MIT, highlights the interest in demystifying the internal processes of these models.

Finding a new antimicrobial compound

In order to refine the selection, several artificial intelligence models were used simultaneously to judge the toxicity of the molecules, compared to three different types of human cells.

AI algorithms scanned nearly 12 million over-the-counter compounds, classifying five families of promising molecules against MRSA. After acquiring and testing nearly 280 of these compounds, two promising candidates were identified.

Positive tests on animal models

Experiments in mice, aimed at simulating either a skin infection or a systemic infection by MRSA, demonstrated the effectiveness of each of the two agents identified, reducing the quantity of bacteria in the treated organisms by ten times.

In short, thanks to the application of artificial intelligence models in the pharmaceutical field, a new era may be dawning for the research and development of antibiotics capable of thwarting the escape strategies of bacteria. resistant.

To read The inevitable technological revolution in robotics: find out why!

Share your opinion