Researchers have used artificial intelligence to make a major breakthrough in countering the world's most dangerous antibiotic-resistant bacteria. In just a few hours, the algorithm was able to find a new super antibiotic, which could destroy the most dangerous antibiotic-resistant bacteria.
Tests have shown that the new antibiotic can kill antibiotic-resistant bacteria, including Acinetobacter and Enterobacteriaceae . These are two of the three levels of the alert level, which the World Health Organization classified as new antibiotic-resistant bacteria.
The new antibiotic can kill most antibiotic resistant bacteria.
This is the first antibiotic found in history. Researcher Regina Barzilay at MIT said: 'I think this is the strongest antibiotic that has been made to date. It can kill some dangerous antibiotic resistant bacteria '.
In order to find a new antibiotic, researchers first have to teach a deep learning algorithm from Google to determine which types of molecules can kill bacteria. To do this, they must provide information on the atomic and molecular features of nearly 2,500 drugs and natural compounds, for AI to study.
After entering enough data, researchers for AI started working to find new antibiotics. However, instead of focusing on finding an effective antibiotic, the algorithm will focus on finding an unprecedented antibiotic. That is the key to destroying the current antibiotic-resistant bacteria.
To expand and search for more powerful new antibiotics, researchers turned to a huge database of about 1.5 billion compounds. They put the algorithm to work with about 107 million of them. After 3 days, AI has come up with a list of 23 new potential antibiotics.
The head of the research project, Dr. Jonathan Stokes, said it was impossible to screen all 107 million compounds manually. Artificial intelligence has helped accomplish this utopia, and shortened the time to just a few days.
Antimicrobial resistance is a concern, without new antibiotics 10 million people around the world could be at risk of infection each year. But with artificial intelligence, this worry will be solved soon.