Photo from open sources
Massachusetts Institute of Technology (USA) established a neural network that has already discovered several new antibiotics.
Researchers have developed a self-learning algorithm that could find 9 drugs that can be used as antibiotics. Scientists are sure that this is a real breakthrough in pharmaceuticals, and humanity has a better chance of survival in case of large-scale epidemics.
According to Cell, the decision to use artificial intelligence to search for new drugs is not new, however, it is the latest development of scientists – a program from MIT made a breakthrough.
The main problem in finding antibiotics is the use of traditional ways. Bacteria most often possess adaptability and very quickly develop protection to antibiotics, which subsequently may be due to uselessness toss out. For example, penicillin, found in 1942, began lose its effectiveness by 1945.
It is currently very difficult to synthesize a new antibiotic chemical method, since 99% of bacteria do not breed in artificial environment, which means that it is impossible to conduct them experiences.
Artificial intelligence, to search for a new drug, was It is proposed to analyze an enormous base of chemical compounds. One of the key parameters was the ability to kill bacteria E. coli. The algorithm immediately discovered 23 substances, among which scientists identified 9 most effective.
According to media reports, one of the new drugs has the name chalitsin – in honor of the artificial intelligence HAL (film “2001”: Space Odyssey “). Chalitsin has a wide spectrum of action, he able to kill tuberculosis bacteria, enterobacteria and akinetobacterium Bauman, pseudomembranous colitis.
Artificial Intelligence
