Particle.news

AI Identifies Thousands of Potential Antibiotics in Archaea

The team will refine structural modelling for archaeasins to support comprehensive safety studies ahead of clinical evaluation.

Overview

  • Researchers retrained the AI tool APEX 1.1 to scan proteomes of 233 archaeal species, uncovering 12,623 candidate antimicrobial peptides dubbed archaeasins.
  • Eighty archaeasins were synthesized and 93% showed in vitro activity against drug-resistant pathogens including E. coli, S. aureus and P. aeruginosa.
  • In animal infection models, three archaeasins arrested a hospital-acquired drug-resistant bacterium within four days, with one matching the efficacy of polymyxin B.
  • Chemical analysis revealed archaeasins differ from known antimicrobial peptides in electric charge distribution and likely disrupt bacterial electrical signalling internally.
  • Authors plan to enhance APEX’s structural prediction capabilities and conduct further efficacy and toxicity testing as they advance toward human trials.