AI Finds 386 Potential Antibiotics in Animal Venoms
To identify new antimicrobial candidates, scientists at the University of Pennsylvania have undermined world venomic data sets using a deep learning system called Apex.
Guan and al. Demonstrate that venoms are a rich source of previously hidden antimicrobial scaffolding, and that the integration of large -scale computer exploitation with experimental validation can speed up the discovery of necessary antibiotics. Image credit: Guan and al., DOI: 10.1038 / S41467-025-60051-6.
The rise of antibiotic resistant pathogens, in particular Gram negative bacteria, highlights the urgent need for new therapies.
Venoms form a huge tank widely unexploited with bioactive molecules with antimicrobial potential.
In the new study, the researcher at the University of Pennsylvania, César de la Fuente, and his colleagues used the APEX system in deep learning to pass through a database of 16,123 venom proteins and 40,626,260 encrypted venom.
Among these, the algorithm has identified 386 candidate peptides which are structurally and functionally distinct from known antimicrobial peptides.
“Venoms are scalable masterpieces, but their antimicrobial potential was barely explored,” said Dr. de la Fuente.
“Apex allows us to scan a huge chemical space in just a few hours and identify peptides with exceptional potential to fight against the most obstinate pathogens in the world.”
From the restricted list selected by AI, scientists synthesized 58 venom peptides for laboratory tests.
Fifty -three of them killed medication resistant bacteria – including Escherichia coli And Staphylococcus aureus – In doses which were harmless to human red blood cells.
“By combining calculation sorting with traditional laboratory experimentation, we have to date one of the most complete research of antibiotics derived from venom,” said Dr. Marcelo Torres, co-author of the study.
“The platform has mapped more than 2,000 entirely new antibacterial patterns-short and specific sequences of amino acids within a protein or peptide responsible for their ability to kill or inhibit bacterial growth,” said Dr. Changge Guan, study co-author.
“Our team now takes the best peptide candidates who could lead to new antibiotics and improve them with medicinal chemistry adjustments.”
The results appear in the newspaper Nature communications.
_____
C. Guan and al. 2025. IT exploration of world venoms for antimicrobial discovery with artificial artificial intelligence. Common nat 16, 6446; DOI: 10.1038 / S41467-025-60051-6
