A study led by César de la Fuente uses AI to discover antibiotics in extinct organisms, such as mammoths. It offers a promising solution to antibiotic resistance.
“Resurrecting” molecules thanks to cutting-edge technologies has become a promising avenue. Now, a New study shows that deep learning can extract proteins from all extinct organisms, also mammoths. The results are published this Tuesday in the journal Nature Biomedical Engineering, in an article led by Spanish César de la Fuente, from the University of Pennsylvania, in the United States.
“Antimicrobial resistance is one of the biggest threats of our time and new antibiotics are urgently needed. Today we report the discovery driven by artificial intelligence (AI) of tens of thousands of -potential- antibiotics in extinct organisms,” summarizes De la Fuente in his X account.
Molecular de-extinction
Molecular de-extinction, powered by machine learning, seeks resurrect molecules from extinct organisms to address problems such as antibiotic resistance. César de la Fuente’s lab at the University of Pennsylvania has developed an AI model called APEX, which has successfully discovered numerous antibiotic compounds in ancient creatures, including the woolly mammoth. This advance is based on decades of research into sequencing ancient genetic material and represents an important step in the emerging field of resurrection biology.
Previously, the lab had explored proteomes of Neanderthals and Denisovans, identifying antibiotic molecules in both modern and ancient humans. To accelerate discovery, APEX was designed to analyze the proteomes of all known extinct organisms. The researchers trained deep learning models that predicted thousands of sequences with broad-spectrum antimicrobial activity, many of which are not found in current organisms, opening new possibilities for the development of innovative antibiotics.
Experiments in mice
The team synthesized 69 peptides and confirmed their activity against bacterial pathogens, observing that most killed bacteria by depolarizing their cytoplasmic membrane. Compounds such as mamuthusin-2 and elefasin-2 showed anti-infective efficacy in mice with skin abscesses or thigh infections, comparable to the antibiotic polymyxin B.
The research, published in Nature Biomedical Engineering, highlights that AI-supported molecular de-extinction can accelerate the discovery of therapeutic molecules. The APEX model has identified promising candidates for preclinical antibiotics, with AI work allowing what previously took years to be done in hours.
With information from Agencia EFE
(Visited 1 times, 1 visits today)
