Reviving Phage Therapy: AI-Powered Solutions for Antibiotic Resistance

**Phage therapy** is experiencing a revival in response to the rise in antibiotic-resistant bacterial infections. Originally sidelined in the 1930s due to the rapid development of antibiotics, this therapy uses bacteriophages or **'phages'** to target and eliminate pathogenic bacteria. Recently, scientists from the Institut Pasteur, Inserm, and Université Paris Cité have developed an **AI model** capable of recommending the best phage combinations for patients by analyzing the genome of the targeted bacteria. This innovation is crucial given the **diversity and specificity** of phages and the growing threat of 'superbugs' like Escherichia coli that resist conventional antibiotics. The AI model was trained on data from 403 E. coli strains and 96 phages, focusing on the interaction mechanisms between bacteria and phages. It analyzes bacterial **membrane receptors**, crucial for phage infectivity, rather than relying on less precise anti-viral defense mechanisms. This precise approach allowed researchers to achieve an **85% accuracy** rate in predicting phage efficacy, exceeding expectations. Further testing on pneumonia-causing E. coli strains demonstrated a **90% success rate** with AI-selected phage cocktails. This AI-driven method not only offers a promising avenue for personalized treatments, but also provides a potential framework adaptable to other bacteria, paving the way for broader applications in combating antibiotic resistance.