AI-Powered Phage Therapy Shows Promise Against Drug-Resistant Infections
A biotech start-up has used an artificial intelligence system to design a personalized bacteriophage treatment for a patient with an E. coli infection resistant to both antibiotics and conventional phage therapies. The AI tool, Omnii-1, successfully targeted 91.4% of 37 unseen E. coli strains using only two phages, outperforming existing prediction models in preliminary testing. However, the results are not peer-reviewed, and due to lack of published clinical outcomes, the treatment’s real-world safety and effectiveness remain unverified.
Radical Numerics, a biotechnology start-up, utilized its AI platform Omnii-1 to develop a personalized bacteriophage therapy for a patient suffering from an E. coli infection that was resistant to both antibiotics and conventional phage treatments. The treatment was administered under the FDA’s compassionate use program, which allows access to experimental therapies in life-threatening cases when no alternatives exist. In laboratory testing, Omnii-1 demonstrated efficacy against 91.4% of 37 previously unseen E. coli strains using only two engineered phages—outperforming two established phage-host prediction models. The results, however, remain unpublished and unreviewed, with the company not disclosing clinical outcomes for the patient. This lack of transparency limits independent verification and raises questions about whether the treatment led to recovery or adverse effects.
The development builds on growing interest in phage therapy as an alternative to antibiotics amid rising antimicrobial resistance. While phages—viruses that infect bacteria—have been studied since the early 20th century, their clinical use has been limited by challenges in identifying effective strains for specific pathogens. Traditional approaches rely on trial-and-error or small libraries of known phages, which often fail against resistant or diverse bacterial strains. Radical Numerics’ AI-driven method aims to overcome this by predicting phage-host interactions with greater precision. The company’
This case highlights a pivotal moment in the evolution of precision medicine for infectious diseases. If validated, AI-powered phage therapy could offer a new tool against drug-resistant infections, which are projected to cause millions of deaths annually by 2050. The approach may disrupt traditional antibiotic development, which has stagnated due to high costs and low returns. Similar to how CRISPR gene editing sparked both excitement and caution in its early applications, AI-designed phage treatments could accelerate innovation while raising ethical and regulatory questions about unproven therapies. The precedent of the 2016 case involving a patient treated with a personalized phage cocktail at the University of California, San Diego—where outcomes were cautiously reported—suggests that such interventions may gain traction if early results are compelling. However, without transparency and independent validation, public trust could erode. What to watch for includes whether other labs