The team tested dozens of peptides for their bacteria-killing abilities in laboratory dishes and selected six potent ones for further testing on mice infected with a common cause of hospital-borne infections. While all six peptides halted the growth of the bacteria, none killed it. The researchers believe that tweaking the molecules and the algorithm could yield more effective results. However, some experts argue that until the algorithm can predict clinically relevant peptides with a higher degree of success, molecular de-extinction may not significantly impact drug discovery.
Key takeaways:
- Bioengineers have used artificial intelligence (AI) to perform molecular ‘de-extinction’, applying computational methods to data about proteins from both modern humans and our long-extinct relatives, Neanderthals and Denisovans.
- This research has allowed the identification of molecules that can kill disease-causing bacteria, potentially inspiring new drugs to treat human infections.
- The researchers trained an AI algorithm to recognize sites on human proteins where they are known to be cut into peptides, and then applied this algorithm to publicly available protein sequences of modern humans, Neanderthals, and Denisovans.
- While the current algorithm didn't yield highly effective molecules, the researchers believe that tweaking the molecules and the algorithm could improve antimicrobial-peptide identification and potentially revolutionize the field of drug discovery.