Researchers at the MIT Institute have recently been able to identify, using an algorithm of artificial intelligence, a very powerful antibiotic compound that is also proving to be extremely effective against bacteria that have already been able to develop antibiotic resistance.
This is a real breakthrough made possible by advances in the world of artificial intelligence that have allowed the computer to scan at high speed, about one hundred million chemical structures and compounds of antibiotics within a few days.
One of the authors of the article on the discovery, Professor James Collins, says: "We wanted to harness the enormous capabilities of artificial intelligence to the field of drug research in order to open a new era in this field of research and indeed we discovered an amazing molecule that is one of the most powerful antibiotics ever discovered."
Scientists in the Laboratory, Archive, Photo: GettyImages
Professor Regina Barzilai of MIT, who co-authored the paper with Prof. Collins, said: "This machine learning model, in which the computer reviews knowledge in the field and discovers new drugs, is very cheap. And a lot of time and money. "
The desire to find new antibiotic drugs has been around for decades because many bacteria have become resistant to existing drugs.
However, this is an expensive journey, as new drugs have usually been found so far on the basis of antibiotics that already exist on the market and have not actually produced new molecules.
This is because of the great difficulty in researching so extensively about the properties of any drug and finding its other combinations that can better deal with the bacteria.
Researchers therefore wanted to take the knowledge of a nervous computer system so that they could examine the field, but according to Prof. Collins "until recently computers did not have the ability to study these structures in depth to identify traits or abilities that would kill bacteria."
As mentioned now, the knowledge already exists and initially the researchers taught the computer to deal with E. coli bacteria.
They taught the computer to scan 2,500 molecules of which 1,700 are FDA-approved drugs and 800 are natural remedies.
antibiotics,
After the computer learned how to analyze these drugs, the researchers uploaded to the computer the information of antibiotics that the institute has, which contains about 6000 drug compounds and the computer processed all the information and found a completely new drug for E. coli bacteria, which does not resemble antibiotics at all.
The new drug was called lecithin, after the artificial intelligence in the space odyssey, and it was found to be very effective against bacteria that have become resistant to other antibiotics while at the same time apparently not harmful to human cells.
In fact - an examination of the lecithin showed that unlike other drugs that the bacteria became resistant to after only 3 days - in this case they did not develop resistance to the drug even after a full month of exposure to it.
It is important to note that the molecule is currently being tested on mice infected with serious diseases to which American soldiers in Iraq have been exposed.
In this test, the drug, which was initially studied against diabetes, showed good treatment in mice and will be studied in humans in the future as well.
After identifying the lecithin, the researchers exposed the computer to the ZINC15 database, an online collection of about 1.5 billion chemical compounds.
Thus within three days the computer found 23 additional unrecognized molecules that are apparently non-toxic to humans and effective against bacteria.
Beyond that, the researchers also plan to test with the software ways to upgrade and improve existing antibiotic drugs to help humanity.
All the information collected will be given to the development of non-profit companies for the benefit of humanity.