Seeking 'Superbug' Antibiotics, Scientists Use AI to Synthesize Molecules from Neanderthals

An anonymous reader shared this report from CNN: The quest for new antibiotics is going back to the Stone Age. The urgency to identify possible candidates has never been greater as the global population faces nearly 5 million deaths every year that are associated with microbial resistance, according to the World Health Organization. A research team led by bioengineering pioneer Cesar de la Fuente is using artificial intelligence-based computational methods to mine genetic information from extinct human relatives such as Neanderthals and long-gone ice age creatures such as the woolly mammoth and giant sloth. The scientists say some of these small protein, or peptide, molecules they have identified have bacteria-fighting powers that may inspire new drugs to fight infections in humans. The innovative work also opens up a completely new way to think about drug discovery. "It has enabled us to uncover new sequences, new types of molecules that we have not previously found in living organisms, expanding the way we think about molecular diversity," said de la Fuente, Presidential Assistant Professor at the University of Pennsylvania, where he heads the machine biology group. "Bacteria from today have never faced those molecules so they may give us a better opportunity at targeting the pathogens that are problematic today...." The idea to look at extinct molecules came up during a lab brainstorm when the blockbuster movie "Jurassic Park" was mentioned. "The notion (in the film) was to bring back entire organisms, and obviously, they had a lot of issues," De la Fuente said. His team started thinking about a more feasible idea: "Why not bring back molecules from the past?" Advances in the recovery of ancient DNA from fossils mean that detailed libraries of genetic information about extinct human relatives and long-lost animals are now publicly available... In research expected to publish next year, de la Fuente and his colleagues have developed a new deep-learning model to explore what he describes as the "extinctome" - the protein sequences of 208 extinct organisms for which detailed genetic information is available. The team found more than 11,000 previously unknown potential antimicrobial peptides unique to extinct organisms and synthesized promising candidates... He said that the peptides they discovered displayed "excellent anti-infective activity" in mice. "Molecular de-extinction offers a unique opportunity to combat antibiotic resistance by resurrecting and tapping into the power of molecules from the past," he said.

Read more of this story at Slashdot.