A new type of compound with antibiotic capabilities has been identified with the help of a novel machine learning approach. The newly identified type of antibiotics, discovered from among more than 100 million molecules, has even one that can work against a range of bacteria like tuberculosis and previously deemed untreatable bacteria.
According to the Termeer Professor of Medical Engineering and Science in MIT’s Institute for Medical Engineering and Science (IMES) and Department of Biological Engineering, James Collins, the team wanted to develop a platform which would help harness the potential of artificial intelligence in regards with antibiotic drug discovery.
He explained, “Our approach revealed this amazing molecule which is arguably one of the more powerful antibiotics that has been discovered.”
The new antibiotic compound called halicin is the first discovered with artificial intelligence. Artificial intelligence or AI refers to the simulation of human intelligence in machines. In this process, machines are programmed to think like humans and mimic their actions. These machines can solve complex problems and calculation within minutes which are otherwise impossible to perform so quickly.
Previously, AI has been used to aid the antibiotic discovery process but this is the first time that it has been used to come up with a new kind of antibiotic from scratch.
Collins and his team screened nearly 107 million molecular structures in a database called ZINC15. From there they narrowed focus to 23 molecules. Further physical tests narrowed it to 8 molecules with antibacterial activity. From there two were identified that were strong enough to defeat a broad range of pathogens. These two could even fight off antibiotic-resistant strains of E.coli as well.
E.coli is a commonly found bacteria in the gut which is usually harmless. However, some of these cause diseases by producing Shiga toxin. Centers for Disease Control and Prevention (CDC) estimates that 265,000 Shiga toxin-producing E. coli (STEC) infections occur each year in the United States.
The research was funded by the Abdul Latif Jameel Clinic for Machine Learning in Health, the Defense Threat Reduction Agency, the Broad Institute, the DARPA Make-It Program, the Canadian Institutes of Health Research, the Canadian Foundation for Innovation, the Canada Research Chairs Program, the Banting Fellowships Program, the Human Frontier Science Program, the Pershing Square Foundation, the Swiss National Science Foundation, a National Institutes of Health Early Investigator Award, the National Science Foundation Graduate Research Fellowship Program, and a gift from Anita and Josh Bekenstein.
Usually antibiotics work by interrupting the activities of bacteria like blocking the enzymes involved in DNA repair, protein synthesis or cell-wall biosynthesis. But halicin works by disrupting the flow of protons across a cell membrane. The antibiotic also showed very low toxicity and was robust against resistance.
Antibiotic resistance happens when bacteria evolves itself in response to medicines. When these bacteria, which have become resistant to antibiotic, infect humans, it becomes next to impossible to treat the disease caused by them with antibiotic which was previously used.
In experiments the resistance usually arise within a day or two, but this did not happen with halicin. Collins said, “Even after 30 days of such testing we didn’t see any resistance against halicin.” This is good news for the scientists working on the subject because currently antibiotic resistance is one of the biggest threats to global health, food security and development today.
Infections such as tuberculosis, pneumonia, gonorrhea and salmonellosis are becoming increasingly harder to treat as antibiotics used against them are becoming less effective. Antibiotic resistance leads to longer hospital stays, higher medical costs and increased death rates.
The computer model used for screening is designed to pick out antibiotic sequences that are the most effective way of killing bacteria. The researchers are planning to use this model for further study to find new antibiotics and optimize the ones that already exist.