According to the Organization for Economic Co-operation and Development (OECD) estimates, antibiotic resistance could cause the death of millions of people by 2050. But a major breakthrough has just been made by researchers at MIT and Harvard. Using the tools of artificial intelligence, they have discovered a new molecule that would be the most powerful antibiotic ever discovered.
The antibiotics currently in use are old, and the traditional process of discovering new antibiotics is costly and burdensome. Artificial intelligence (AI) allows us to search “in silico”, i.e., through computer modeling, for chemical molecules that could attack certain bacteria.
We wanted to develop a platform to harness the power of artificial intelligence and begin a new era of antibiotic discovery,” said James Collins, professor of medical engineering at the Massachusetts Institute of Technology (MIT) and co-author of the discovery, which was published in the Peer-reviewed journal Cell. Our approach discovered this incredible molecule, which is arguably the most powerful antibiotic ever discovered.
A long overdue breakthrough
AI allows the scope of drug candidates to be expanded to include molecules that researchers had not previously suspected. The idea has been around for decades, but so far, the methods have not been mature enough to find really effective molecules. The researchers trained their model using Escherichia coli bacteria and then searched a library of 6,000 chemical compounds that had the desired properties. The algorithm found a compound with a different structure from existing antibiotics and predicted that it would be effective against many bacteria.
They named the molecule “Halicin” after the HAL computer in the 2001 “Space Odyssey” film, and then tested it in a laboratory against dozens of bacterial strains taken from patients and cultivated in vitro. Halicin has successfully killed many available antibiotic-resistant bacteria, including Clostridium difficile, Acinetobacter baumannii, and Mycobacterium tuberculosis. Only Pseudomonas aeruginosa bacteria are resistant to Halicin.
Finally, the new molecule was tested in rats infected with A. baumannii, a bacterium that infected many American soldiers in Iraq and Afghanistan and that is resistant to all existing antibiotics. The rats were cured within 24 hours.
The authors hope their model will help strengthen the entire antibiotic arsenal at a time when antibiotic resistance is a global concern for health authorities. The Organization for Economic Cooperation and Development (OECD) has recently estimated that resistant bacteria could kill 2.4 million people in Europe, North America, and Australia by 2050.
References
Stokes, J. M., Yang, K., Swanson, K., Jin, W., Cubillos-Ruiz, A., Donghia, N. M., MacNair, C. R., French, S., Carfrae, L. A., Bloom-Ackermann, Z., Tran, V. M., Chiappino-Pepe, A., Badran, A. H., Andrews, I. W., Chory, E. J., Church, G. M., Brown, E. D., Jaakkola, T. S., Barzilay, R., & Collins, J. J. (2020). A deep learning approach to antibiotic discovery. Cell, 180(4), 688–702.e13. https://doi.org/10.1016/j.cell.2020.01.021
Trafton, A. (2020, February 20). Artificial intelligence yields new antibiotic. MIT News. https://news.mit.edu/2020/artificial-intelligence-identifies-new-antibiotic-0220
