Researchers from the Massachusetts Institute of Technology (MIT), Brigham and Women’s Hospital, and Duke University have used tissue models and machine-learning algorithms to identify how specific drugs pass through the digestive tract. The knowledge can help improve patient treatments, as certain drugs could interfere with each other if they depend on the same protein transporters. The team identified that a prescribed antibiotic and a blood thinner interfere with each other. Drug-makers could leverage these findings to create drugs that are more easily absorbed by adding excipients to enhance interactions with transporters.
The team used open-source databases coupled with a customized in vitro tissue model that gauges drug absorption to create a dataset to train a machine-learning model to identify potential interactions in drugs. Following the training, the program made nearly 2 million predictions on how various drugs interact with the GI tract and assessed the accuracy of the predictions.
Among the predictions, the machine-learning model identified that an antibiotic called doxycycline could potentially interact with warfarin, a blood thinner; digoxin, a heart failure medication; levetiracetam, an antiseizure medication,; and tacrolimus, an immunosuppressant. The interaction between doxycycline and warfarin was confirmed using data from about 50 patients who were taking one of those three drugs when the antibiotic was prescribed. The patients’ blood levels of warfarin notably increased when they started taking doxycycline and decreased after they stopped taking the antibiotic. This feedback confirmed the absorption of doxycycline is affected by digoxin, levetiracetam, and tacrolimus.
In addition to recognizing potential interactions between drugs, this approach could facilitate the development of new drugs. This technology is now being used by the biotech company Vivtex to optimize the formulation of new drug molecules to either prevent interactions with other drugs or improve their absorbability.
The study, funded partly by the U.S. National Institutes of Health, the Department of Mechanical Engineering at MIT, and the Division of Gastroenterology at Brigham and Women’s Hospital, was recently published in Nature Biomedical Engineering.
