Artificial intelligence (AI) is making game-changing achievements in every sphere and recently the use of AI has proved to be a game changer in the health sector. Artificial intelligence now has helped scientists to create a new antibiotic in 60 years.
James Collins, professor of Medical Engineering and Science at the Massachusetts Institute of Technology (MIT) and one of the study’s authors, said in a statement: “The insight here was that we could see what was being learned by the models to make their predictions that certain molecules would make for good antibiotics.”
The arrival of a new compound that can destroy drug-resistant bacterium, which caused Thornlands of deaths every year, however, is a milestone in the health industry.
“Our work provides a framework that is time-efficient, resource-efficient, and mechanistically insightful, from a chemical-structure standpoint, in ways that we haven’t had to date,” the statement added.
To make the selection process of potential drugs more efficient and refined, the researchers used deep-learning models. The learning models were trained to calculate the toxicity of compounds in three different types of human cells. After that, by integrating these toxicity assessment results with the pre-determined antimicrobial activity, the researchers identified compounds that can efficiently combat microbes with minimal harm to the human body.
The use of the deep-learning models facilitated the screening of around 12 million commercially available compounds. The models located compounds from five different classes, classified based on special chemical substructures within the molecules, that showed forecasted activity against MRSA. As a result, the scientists got 280 compounds and carried out tests against MRSA in a laboratory setting, helping the researchers to recognize two promising antibiotic candidates from the same category.
However, the deep-learning models used to predict the activity and toxicity of the new compound involved the use of artificial neural networks to automatically learn and represent features from data without explicit programming.
The Massachusetts Institute of Technology team trained an extensively enlarged deep learning model using expanded datasets and a postdoc in the team of researchers said: “What we set out to do in this study was to open the black box. These models consist of very large numbers of calculations that mimic neural connections, and no one really knows what”s going on underneath the hood.”
However, the first new antibiotic discovery in over 60 years using AI has set a precedent for the future leap in the sector.