Artificial Intelligence has already proven to be a boon for several industries, and the pharmaceutical industry is no exception to that.
Fremont, CA: The past few decades have seen how technology is being deployed for the study and creation of new drugs. Using AI or predictive computer models have not yielded accurate results in the past, hence this new model could be a breakthrough for the pharma sector. Currently, newer neural networks are able to pick up on representations easily and map the molecules into continuous vectors that will be used to predict its properties.
Recently, a team of researchers at the Massachusetts Institute of Technology, with the help of a machine learning (ML) algorithm, has found a new antibiotic compound. This new drug is powerful enough to kill several disease-causing bacteria is also found to have killed some strains that have been resistant to all other antibiotics. The drug also helped clear infections in two independent mouse models.
The ML algorithm is capable of screening more than millions of chemical compounds within a minimal amount of time. This model has been developed to select potential and powerful antibiotics that can kill bacteria through the several mechanisms of the existing antibiotic drugs. The research team aimed at developing a platform that will enable them to leverage the power of artificial intelligence. Their approach helped them identify one of the most powerful molecules.
The ML algorithm also helped the research team identify other antibiotic candidates that could prove to be useful. The team believes that this model can be used to create new antibiotics as well.
In this experiment, the researchers designed the model to identify chemical features that make molecules powerful enough to kill E. coli. After the model was trained, it was tested on the Broad Institute's Drug Repurposing Hub, which is a library of about 6,000 compounds. Out of them, the model identified the strongest molecule with unique chemical structures. The research team also designed another ML model to prove that the antibiotic will have a low toxicity rate on human cells.