← Back to Home

AI-Designed Phages Administered to Patient with Drug-Resistant Infection; Efficacy and Safety Results Not Yet Public

Radical Numerics announced that a formulation containing AI-designed phages has been administered to a patient through compassionate use. Moving from the laboratory into human use marks progress, but public clinical evidence of whether it can control infection is still lacking.

By SURL BioNews

When antibiotics struggle to control an infection, finding bacteriophages that can kill the disease-causing bacteria may also be a race against time. Radical Numerics announced on October 20 that phages designed by its Omnii-1 model had been administered to a patient with a drug-resistant bacterial infection through an authorized, physician-led compassionate use pathway. This brings a therapy designed with AI involvement into human use, but there is currently no public answer as to whether the patient benefited.

According to the company’s press release issued through Business Wire, the research team designed phages targeting the bacterial strain causing the patient’s infection within weeks. A mixed formulation containing AI-designed phages first killed bacterial isolates from the patient in laboratory tests before being administered to the patient. This result supports the formulation’s antibacterial activity in vitro, but the performance of the mixture alone cannot establish how much each AI-designed component contributed.

Bacteriophages are viruses that infect bacteria. They must first recognize the bacterial surface, then complete infection and replication to exert their bacteria-killing effects. This specificity offers the potential to target disease-causing bacteria precisely, while also creating a matching challenge: a phage that kills one bacterial strain may not work against bacteria in another patient. AI’s specific task here is to help design and screen candidates that are more likely to match.

The scope of “AI-designed” also needs clarification. The company’s technical description involves modifying specific regions of phage genomes, including receptor-binding proteins that affect bacterial recognition, while retaining the rest of the genome’s architecture. This does not establish that the entire genomes of the phages used in this patient were generated from scratch. The report uses the design of T7 phage tail fiber proteins as an experimental example, but these platform experiments are not equivalent to patient treatment outcomes.

Omnii-1 is trained on genomic and multi-omics data and uses publicly available phage–bacteria interaction data to learn matching relationships, then uses experimentally measured activity to help rank candidate designs. This process still depends on experimental validation: model predictions can narrow the search, but cannot directly guarantee that candidate phages will be effective in the environment of an infection in the human body.

The company describes this administration as the world’s first human use of AI-designed phages. That claim currently comes primarily from the company’s announcement and its press release, neither of which constitutes independent clinical validation. Citing privacy protection, the company has not disclosed the patient’s clinical course or outcomes, so it remains impossible to assess whether the infection improved, how long any effects lasted, or whether adverse reactions occurred.

Compassionate use gives seriously ill patients who lack suitable treatment options an opportunity to receive investigational therapies outside clinical trials; it does not mean a product has been approved for general use. This announcement reports progress from design and in vitro testing to actual administration. Assessing whether it can become a reliable tool for personalized infection treatment will require clinical data available for review, along with more systematic studies of safety and efficacy.

References

  1. 조선일보
  2. Radical Numerics
  3. Radical Numerics via Business Wire / Times-Herald