Phagos, a Paris‑based biotech startup, is deploying an artificial‑intelligence model built on Amazon Web Services to accurately match bacteriophages with antibiotic‑resistant bacterial strains. This development marks a pivotal shift from arbitrary trial‑and‑error methods to precision medicine in combating infections that claim an estimated one million lives annually.
The challenge of antibiotic resistance has spurred Phagos, co‑founded by microbiologist Adèle James and economist Alexandros Pantalis, to pursue alternative therapies. Phages are viruses that specifically infect bacteria, a property that enables targeted treatment without disrupting beneficial microorganisms. But identifying the right phage for a given pathogen has traditionally required labour‑intensive screening of millions of combinations.
Phagos is leveraging AWS’s Generative AI Accelerator Programme, which has provided them with cloud credits, embedded AWS architects and machine‑learning engineers, and dedicated mentorship, to build a generative AI platform capable of predicting effective pairings between phages and bacterial strains. The AI model has been trained on datasets from public repositories and proprietary lab results.
“We can now read into the DNA and see if there’s going to be an interaction between a phage and a bacterium,” Pantalis explained. This capability bypasses exhaustive wet‑lab testing and accelerates the discovery pipeline.
Phagos’s strategy involves initial trials in animal health—shrimp, chicken, bovine, swine—with phages administered through water systems to curb infections in farm environments. James noted that reducing antibiotic use in livestock is critical, as drug‑resistant bacteria generated in farms can transfer to humans via environmental pathways.
The startup’s first validation came from a collaboration with an oyster farmer in France, where AI‑matched phages slashed mortality by 40 per cent. Technical infrastructure from AWS played a crucial role, with cloud experts helping to refine the system and scale it efficiently. “We became much faster, and spent way less time figuring out how to set up our infrastructure,” James said.
Globally, the antibiotic resistance crisis is projected to cause more deaths than cancer by mid‑century, with economic impacts exceeding US$1.5 trillion. By automating phage matching, Phagos aims to create personalised and adaptive therapies that evolve in tandem with bacterial mutations.
Participation in AWS’s 2024 accelerator—which selected just 80 startups from over 4,700 applicants—has provided Phagos not only with technical and financial support, but also exposure to industry networks and potential funding sources.
The platform-as‑a‑service model being developed by Phagos positions the company as a pioneer in AI‑driven microbiological model building. One quarter of its team now focuses exclusively on data and machine learning, a rarity in biotech.
Expansion plans are underway: the company intends to move into human therapeutic applications by 2030, while continuing its work in agriculture. It is also advancing bespoke phage therapies, distinguishing itself from generic phage cocktails pursued by competitors.
While obstacles remain—such as regulatory pathways for phage-based treatments and the complexity of bacterial ecosystems—Phagos believes its AI model and cloud infrastructure provide a scalable solution. Each new phage cocktail generates data that feeds into the platform, enhancing future predictions.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
