Researchers have discovered two new possible antibiotics created by artificial intelligence that may be able to eradicate MRSA and drug-resistant gonorrhea.
The AI created the medications atom by atom, and in both lab and animal tests, they eradicated the superbugs.
Before being given, the two chemicals still require years of clinical studies and development.
However, its creators at the Massachusetts Institute of Technology (MIT) claim AI might usher in a "second golden age" in the development of antibiotics.
Although germs are killed by antibiotics, diseases that are resistant to therapy are now responsible for over a million fatalities annually.
Bacteria have evolved to evade the effects of antibiotics due to overuse, and there has been a decades-long scarcity of novel antibiotics.
In the past, scientists have employed AI to search through thousands of known compounds to find compounds that might develop into novel antibiotics.
The MIT team has now taken things a step further by employing generative AI to initially build drugs for the potentially fatal MRSA (methicillin-resistant Staphylococcus aureus) and the sexually transmitted disease gonorrhea.
36 million molecules, including ones that are either nonexistent or have not yet been found, were examined in their study, which was published in the journal Cell.
In order to train the AI, scientists provided it with information on the chemical structures of well-known substances as well as whether or not they inhibit the growth of various bacterial species.
The AI then discovers how various molecular structures—which are composed of atoms like carbon, oxygen, hydrogen, and nitrogen—affect bacteria.
Then, two methods for using AI to create novel antibiotics were attempted. By looking through a library of millions of chemical fragments, ranging in size from eight to nineteen atoms, the first discovered a good beginning place and proceeded to build from there. The second gave the AI complete control right away.
Additionally, anything that resembled existing antibiotics too much was eliminated during the design phase. Additionally, it made an effort to weed out anything that was anticipated to be harmful to people and make sure they were creating medications rather than soap.
Antibiotics for gonorrhea and MRSA, a form of bacteria that lives harmlessly on the skin but can cause a deadly infection if it enters the body, were developed by scientists using artificial intelligence.
Two new possible medications were produced after the top designs were tested on infected mice and bacteria in a lab.
"We're excited because we show that generative AI can be used to design completely new antibiotics," MIT professor James Collins tells the BBC.
"AI can enable us to come up with molecules, cheaply and quickly, and in this way, expand our arsenal, and really give us a leg up in the battle of our wits against the genes of superbugs."
However, they are not yet ready for clinical trials, and the medications will need to be refined, which is expected to take another one to two years, before the arduous process of testing them on humans can begin.
Dr. Andrew Edwards of the Fleming Initiative and Imperial College London described the discovery as "very significant" with "enormous potential" since it "demonstrates a novel approach to identifying new antibiotics".
He did say, however: "While AI promises to dramatically improve drug discovery and development, we still need to do the hard yards when it comes to testing safety and efficacy."
This can be a lengthy and expensive process, with no certainty that the experimental drugs will be provided to patients in the end.
Some advocate for broader improvements in AI drug discovery. Prof Collins says "we need better models" that go beyond how well medications perform in the lab to better anticipate their usefulness in the body.
There is also a concern about how difficult the AI designs are to construct. Only two of the top 80 gonorrhea therapies suggested in theory have been synthesised into medications.
Prof Chris Dowson of the University of Warwick said the study was "cool" and demonstrated AI was a "significant step forward as a tool for antibiotic discovery to mitigate against the emergence of resistance".
However, he notes that there is an economic issue that contributes to drug-resistant infections: "How do you make drugs that have no commercial value?"
If a new antibiotic were produced, it would ideally be used as infrequently as possible to retain its potency, making it difficult for anyone to benefit.
0 Comments