Bioengineers have brought ancient molecules back from the dead using artificial intelligence.

 

Resurrecting Extinct Antibiotics

The researchers carried out the process of molecular 'de-extinction' by applying computational methods to data about proteins taken from modern humans (Homo sapiens) and our extinct relatives, Neanderthals (Homo neanderthalensis) and Denisovans. The study allowed them to identify molecules that can kill disease-causing bacteria and gain insight into developing new drugs.

According to bioengineer Cesar de la Fuente from the University of Pennsylvania in Philadelphia, their research team is motivated by bringing back ancient molecules to address the current problems encountered by humans. They were inspired by the classic blockbuster film "Jurassic Park" and thought they could bring back ancient molecules instead of resurrecting dinosaurs.

To make this possible, the scientists trained an AI algorithm to help them spot the sites on human proteins where they are divided into peptides. The algorithm was applied to publicly available protein sequences to detect the new peptides. Then the researchers used the behavior of previously described antimicrobial peptides in predicting the unique peptides that can kill bacteria.

The researchers tested dozens of peptides to find out if they could kill bacteria in a laboratory setting. Six potent peptides were selected - four from Homo sapiens, one from Homo neanderthalensis, and one from Denisovans. These peptides were given to mice nfected with Acinetobacter baumannii, a tacterium that causes hospital-borne infections in humans.

It was found that the growth of A. baumannii in the thigh muscles of mice was stopped by the six peptides, but none of them killed the bacteria. Five of the molecules killed the bacteria that thrive in skin abscesses in extremely high doses.

From their study, the researchers realized that adjusting the most successful molecules can help create more practical versions. On the other hand, modifying the algorithm can enhance antimicrobial-peptide identification with less false positives. Even if the algorithm the researchers used does not yield amazing molecules, the researchers believe its framework can provide a new avenue for improved drug discovery.

It takes three to six years to discover a new antibiotic using older drug development methods. Meanwhile, it only takes weeks to find and test drug candidates using AI.

READ ALSO: Antibiotic-Resistance Microbes Had Predecessor 450-Million Years Ago Before The Mesozoic Era

The Need for Improved Antibiotic Development

Since the discovery of penicillin by Alexander Fleming in 1928, it took another decade before this substance was introduced as a potential treatment for bacterial infections. The development of antibiotics has slowed over the past decades, and most antibiotics currently prescribed have existed in the market for over 30 years. Meanwhile, misuse of this drug has given rise to antibiotic-resistant bacteria, so experts are trying to look for a new wave of treatments.

Most organisms produce short protein subunits known as peptides which are known for their antimicrobial properties. Many antimicrobial peptides isolated from bacteria are already in clinical use. Experts believe that proteins from extinct species can show potential in antibiotic development. The revival of the extinct antibiotic molecules encoded in the DNA of Neanderthals and Denisovans could provide a key to fighting against antimicrobial resistance.

 

RELATED ARTICLE: How Do Antibiotic-Resistant Germs Bring Toxins to Other Cells? Scientists Explain

Check out more news and information on Antibiotics in Science Times.