With the help of artificial intelligence (AI), scientists have developed an antibiotic that can destroy a deadly type of superbug. UI quickly reviewed and greatly reduced the list of several thousand potentially active substances, which were then tested in the laboratory.
The result of the tests is a powerful, experimental antibiotic abaucinwhich will have to be further tested before general use in medicine, reports the BBC.
As is known, antibiotics destroy bacteria, but scientists have not developed new drugs for decades. Bacteria are therefore increasingly difficult to overcome, as they have become resistant to the antibiotics that medicine has at its disposal. According to experts, more than a million people die each year from infections with bacteria that are resistant to antibiotic treatment.
In this research, the researchers focused on one of the most problematic types of bacteria, Acinetobacter baumannii, which can infect wounds and cause pneumonia. It is one of the three superbugs designated by the World Health Organization (WHO). “critical” a threat.
The bacterium is often resistant to multiple antibiotics and is a major problem in hospitals and nursing homes as it is persistent and can survive on surfaces and medical equipment.
UI coming to the rescue?
Researchers in Canada and the US say artificial intelligence can greatly speed up the development of new drugs.
For said research, they first had to train the artificial intelligence and teach it what chemicals/substances could be effective. So, thousands of drugs whose exact chemical structure is known were first manually tested on said bacteria to see which would slow or kill it. This data was then fed into an artificial intelligence program to “learn” the chemical characteristics of drugs that might attack the problematic bacteria.
In the next phase, artificial intelligence examined 6,600 ingredients, the effectiveness of which is currently unknown to scientists. It only took the program an hour and a half to work its way through the “list” and the result was a shorter list of potential agents.
The researchers then tested 240 chemicals from this list in the laboratory and discovered nine potential new antibiotics among them. One of them is incredibly powerful abaucinwhich lab experiments have shown to heal infected wounds in mice and kill the bacteria Acinetobacter baumannii.
Refinement and clinical trials follow
The next step is to perfect the drug in the laboratory, followed by clinical trials. The first antibiotics developed with the help of artificial intelligence are expected to be available in 2030.
Interestingly, the mentioned experimental antibiotic only worked against the bacteria Acinetobacter baumannii, but had no effect on other species. It is this precision of abaucin that could mean that resistance to it will be more difficult to develop, and there could also be fewer side effects of the treatment.