MIT Researchers Leverage AI to Uncover Promising Class of Antibiotic Candidates
Using a type of artificial intelligence known as deep learning, MIT researchers have made a breakthrough discovery in the fight against drug-resistant bacteria. They have identified a class of compounds that can effectively kill methicillin-resistant Staphylococcus aureus (MRSA), a bacterium responsible for thousands of deaths in the United States each year.
In a study published in Nature, the researchers demonstrated that these compounds exhibited potent antimicrobial activity against MRSA in laboratory experiments and two mouse models of MRSA infection. Importantly, the compounds showed minimal toxicity to human cells, making them strong candidates for future drug development.
One key innovation of the study was the ability to determine the information that the deep learning model used to generate its antibiotic potency predictions. This knowledge could aid researchers in designing even more effective drugs based on the insights gained from the model.
James Collins, the Termeer Professor of Medical Engineering and Science at MIT, explained the significance of this insight, stating, Our work provides a framework that is time-efficient, resource-efficient, and mechanistically insightful… in ways that we haven’t had to date.
The research project, part of the Antibiotics-AI Project at MIT led by Collins, aims to discover new classes of antibiotics against seven types of deadly bacteria within a seven-year timeframe.
MRSA is a highly prevalent bacterium in the United States, causing thousands of infections annually. It commonly results in skin infections, pneumonia, and in severe cases, sepsis, which can be fatal.
Collins and his colleagues at the Abdul Latif Jameel Clinic for Machine Learning in Health have been utilizing deep learning to uncover potential new antibiotics. Their previous work successfully identified potential drugs against Acinetobacter baumannii and other drug-resistant bacteria.
However, a limitation of their approach was the lack of transparency in the models’ predictions, also known as black boxes. The researchers sought to address this limitation in the current study by developing an algorithm that provided information about both the antimicrobial activity of the compounds and the specific substructures within the molecules responsible for this activity.
To identify the most promising drug candidates, the researchers trained additional deep learning models to predict the toxicity of the compounds to different types of human cells. By combining this toxicity data with the antimicrobial predictions, they narrowed down a pool of approximately 12 million commercially available compounds to five classes that exhibited activity against MRSA.
After purchasing and testing around 280 of these compounds, two compounds from the same class emerged as highly effective antibiotic candidates. In mouse models, these compounds reduced the MRSA population by a factor of 10. Further experiments suggested that the compounds work by disrupting bacteria’s ability to maintain an electrochemical gradient across their cell membranes, ultimately leading to their destruction.
The researchers have shared their findings with Phare Bio, a nonprofit organization established by Collins and others as part of the Antibiotics-AI Project. Phare Bio will conduct further analysis to evaluate the chemical properties and potential clinical use of these compounds.
Meanwhile, Collins’ lab is already focused on designing additional drug candidates based on the study’s findings. Their deep learning models will also continue to search for compounds with the potential to combat other types of bacteria.
This groundbreaking research represents a significant step forward in the development of new antibiotics and provides hope for addressing the global challenge of antibiotic resistance. By leveraging artificial intelligence, MIT researchers have identified a promising class of compounds that may pave the way for more effective treatments against drug-resistant bacteria, ultimately saving lives and improving healthcare outcomes for patients worldwide.
