AI Creates Novel Antibiotics Against Gonorrhea & Staph Infections
A team of researchers at Massachusetts Institute of Technology (MIT) used artificial intelligence of generative types to get two new antibiotic potentialStaphilococcus aureus meticillin resistant).
The drugs were designed atom with atom by Ai and killed the superbacteries in laboratory tests and animals. Of course, the two compounds also need additional tests, years of improvement and clinical studies before they are proposed to patients.
Why antibiotic resistance is a global problem
The team from MIT commented that this discovery could mark the beginning of a “second golden age” in the discovery of antibiotics, in a world where such infections cause over one million deaths a year.
Due to the excessive use of antibiotics, the bacteria have evolved, and the last decades have been marked by an acute lack of new antibiotics.
The fact that generative artificial intelligence has created these products is a huge step for research in this field.
Staphilococcus aureus is harmless if it is on the skin, but once entered the body through wounds, cuts, etc., it is devastating.
How did he get to create drugs
Also, gonorrheacaused by the Bacteria Neisseria Gonorrhoeae, which has developed resistance over time, in all classes of antibiotics and which WHO has classified “priority pathogen”, develops, in some cases, a super-tulp, which is extremely difficult to cure and which does not respond to the standard treatments.
Scientists have trained the AI by giving them the chemical structure of well-known compounds, together with data on the extent to which they slow down the growth of different species of bacteria.
Thus, you have learned how bacteria of different molecular structures are affected, made up of atoms such as carbon, oxygen, hydrogen and nitrogen.
Two approaches were tested for designing new antibiotics using AI.
The first identified a promising starting point looking in a library of millions of chemical fragments, of eight to 19 atoms, and built from there.
The second offered the total freedom from the beginning.
The design process also eliminated any compound too similar to existing antibiotics.
There’s more so far but the beginning is promising
It was also tried to guarantee the fact that drugs are created, not substances such as soap, and filtering any compound provided to be toxic to humans.
After manufacture, the most promising models were tested on bacteria in the laboratory and infected mice, resulting in two new potential drugs.
“We are excited because we have shown that generative artificial intelligence can be used to create completely new antibiotics,” said Professor Collins, from MIT, for the BBC.
“AI can help us conceive cheap and fast molecules and, in this way, expand our arsenal and give us an advantage in our intelligence against the genes.”
However, they are not prepared for clinical tests, and drugs will require refining – estimated in one more or two years of work – before the long testing process can start.