Artificial intelligence is transforming the search for new treatments for Parkinson's disease, antibiotic-resistant superbugs and rare diseases, offering hope for conditions long considered incurable. Scientists at the Massachusetts Institute of Technology (MIT) have used AI to discover two new antibiotic compounds effective against highly drug-resistant gonorrhoea and MRSA, marking a significant breakthrough in the fight against antimicrobial resistance.
Bacterial resistance to antibiotics is a growing global crisis, with around 1.1 million deaths annually from previously treatable infections. The death toll is projected to exceed eight million by 2050 without urgent action. Between 2017 and 2022, only 12 new antibiotics were approved, most similar to existing drugs. The field has suffered from underfunding and lack of pharmaceutical interest.
Professor James Collins of MIT and his team trained a generative AI model to recognise chemical structures of known antibiotics. The AI screened over 45 million compounds, identifying two that kill drug-resistant strains of Neisseria gonorrhoeae and Staphylococcus aureus (MRSA). These compounds appear to work differently from existing antibiotics, potentially forming a new class of medicines.
The AI designed 36 million potential compounds, of which 24 were synthesised in the lab. Seven showed antimicrobial activity, and two were highly effective against resistant bacteria. The candidates are now undergoing further testing. Collins' lab previously used AI to discover antibiotics effective against Clostridium difficile and Mycobacterium tuberculosis.
For diseases like Parkinson's, where no treatment slows progression, AI offers new hope. With over 10 million patients worldwide and rising rates in ageing populations, researchers are applying AI to conditions with limited existing knowledge, aiming to unlock breakthroughs where traditional methods have failed.
