Scientist Find Most Powerful Antibiotic Through AI, It Kills Most Drug-Resistant Bacteria

Antibiotics are our only hope against bacterial infections. However, some kinds of bacteria are getting resistant to antibiotics, requiring the need for stronger alternatives. And now researchers at MIT have discovered one with the help of artificial intelligence.

Reuters

This is the first-ever bacteria that has been discovered using AI that has gone through extensive digital libraries of pharmaceutical compounds.

Published in the journal Cell, the newly discovered antibiotic has managed to wipe out a range of antibiotic-resistant strains of bacteria including Acinetobacter baumannii and Enterobacteriaceae, which according to WHO two of three high-priority pathogens, ranked as critical for new antibiotics to target.

Bacteria become resistant to an antibiotic when it mutates and evolves to circumvent mechanisms used by antibiotics to kill them. Thus arose the need to find new antibiotics to kill the powerful bacteria.

For finding the super antibiotic, researchers first trained a deep-learning algorithm to help it recognise the kind of molecules that kill bacteria. They fed the program information on the atomic and molecular features of around 2,500 drugs and natural compounds and the performance of those substances against the deadly E Coli.

Once the algorithm understood what they wanted for the next antibiotic, scientists then let it go through a library of over 6000 compounds under investigation for treating numerous human diseases.

Here instead of looking for antimicrobials, the AI was focussing on compounds that seemed effective but nowhere close to what was currently being commonly used as antibiotics. This basically enabled a never-before-seen compound that the bacteria had never been exposed to and by extension not develop resistance for.

Reuters

Within hours, AI was giving out promising antibiotics -- one of them was called ¡®Halicin¡¯ based after Hal from the movie 2001: A Space Odyssey. Post testing they found that it had the ability to kill

Mycobacterium tuberculosis, that causes TB. It also affected strains of Enterobacteriaceae that are resistant to carbapenems -- a set of antibiotics that¡¯s a last resort for infections like these. Halicin also killed C difficile and multidrug-resistant Acinetobacter baumannii infections in mice.

To look for more drugs, they let teh AI go through a massive database of 1.5 billion compounds with the algorithm prepped with 107 million of these. In just three days it delivered 23 potential antibiotics.

Scientists are on the lookout for more compounds in other databases. Additionally, they also want to look for a selective antibiotic that only kills specific harmful bacteria and not healthy ones in our body.