However de la Fuente is utilizing synthetic intelligence to carry a couple of totally different future. His crew on the College of Pennsylvania is coaching AI instruments to go looking genomes far and deep for peptides with antibiotic properties. His imaginative and prescient is to assemble these peptides—molecules fabricated from as much as 50 amino acids linked collectively—into numerous configurations, together with some by no means seen in nature. The outcomes, he hopes, may defend the physique towards microbes that stand up to conventional therapies.
His quest has unearthed promising candidates in sudden locations. In August 2025 his crew, which incorporates 16 scientists in Penn’s Machine Biology Group, described peptides hiding within the genetic code of historic single-celled organisms referred to as archaea. Earlier than that, they’d excavated an inventory of candidates from the venom of snakes, wasps, and spiders. And in an ongoing venture de la Fuente calls “molecular de-extinction,” he and his collaborators have been scanning printed genetic sequences of extinct species for doubtlessly purposeful molecules. These species embody hominids like Neanderthals and Denisovans and charismatic megafauna like woolly mammoths, in addition to historic zebras and penguins. Within the historical past of life on Earth, de la Fuente causes, possibly some organism advanced an antimicrobial protection that may very well be useful at this time. These long-gone codes have given rise to resurrected compounds with names like mammuthusin-2 (from woolly mammoth DNA), mylodonin-2 (from the enormous sloth), and hydrodamin-1 (from the traditional sea cow). Over the previous few years, this molecular binge has enabled de la Fuente to amass a library of greater than one million genetic recipes.
At 40 years previous, de la Fuente has additionally collected a trophy case of awards from the American Society for Microbiology, the American Chemical Society, and different organizations. (In 2019, this journal named him one in all “35 Innovators Underneath 35” for bringing computational approaches to antibiotic discovery.) He’s well known as a pacesetter within the effort to harness AI for real-world issues. “He’s actually helped pioneer that area,” says Collins, who’s at MIT. (The 2 haven’t collaborated within the laboratory, however Collins has lengthy been on the forefront of utilizing AI for drug discovery, together with the seek for antibiotics. In 2020, Collins’s crew used an AI mannequin to foretell a broad-spectrum antibiotic, halicin, that’s now in preclinical growth.)
The world of antibiotic growth wants as a lot creativity and innovation as researchers can muster, says Collins. And de la Fuente’s work on peptides has pushed the sphere ahead: “César is marvelously proficient, very progressive.”
A messy, noisy endeavor
De la Fuente describes antimicrobial resistance as an “nearly not possible” drawback, however he sees loads of room for exploration within the phrase nearly. “I like challenges,” he says, “and I believe that is the last word problem.”
The use, overuse, and misuse of antibiotics, he says, drives antimicrobial resistance. And the issue is rising unchecked as a result of typical methods to search out, make, and check the medication are prohibitively costly and sometimes result in useless ends. “Plenty of the businesses which have tried to do antibiotic growth prior to now have ended up folding as a result of there’s no good return on funding on the finish of the day,” he says.
Antibiotic discovery has at all times been a messy, noisy endeavor, pushed by serendipity and fraught with uncertainty and misdirection. For many years, researchers have largely relied on brute-force mechanical strategies. “Scientists dig into soil, they dig into water,” says de la Fuente. “After which from that advanced natural matter they attempt to extract antimicrobial molecules.”
However molecules might be terribly advanced. Researchers have estimated the variety of potential natural mixtures that may very well be synthesized at someplace round 1060. For reference, Earth incorporates an estimated 1018 grains of sand. “Drug discovery in any area is a statistics recreation,” says Jonathan Stokes, a chemical biologist at McMaster College in Canada, who has been utilizing generative AI to design potential new antibiotics that may be synthesized in a lab, and who labored with Collins on halicin. “You want sufficient photographs on purpose to occur to get one.”
