Researchers Use AI To Learn Which Drugs Don’t Mix

the study, It was published in the magazine natural biomedical engineeringcenters around a model used to determine which drugs may interfere with each other when taken together.

“An in vitro system that accurately models the in vivo conditions of the gastrointestinal tract could aid in the development of more bioavailable oral drugs,” the researchers wrote.

“Here we show that the interaction profile between drugs and intestinal drug transporters is altered by modulating transporter expression in intact porcine tissue explants via ultrasound-mediated delivery of small interfering RNA. We show that the interaction profiles obtained can be classified by a trained random forest model.'' on drug-transporter relationships. ”

According to MIT News, According to a paper describing the study, the researchers “utilized both tissue models and machine learning algorithms” to “confirm that commonly prescribed antibiotics and blood thinners can interfere with each other. It became clear.”

The magazine says, “Further understanding of which transporters help drugs pass through the gastrointestinal tract will help drug developers improve the absorption of new drugs by adding excipients that enhance interactions with transporters.” It may also be helpful to do so.” Similarly, it may be “applicable to drugs currently under development.”

“Using this technology, drug developers can tailor the formulation of new drug molecules to prevent interactions with other drugs or improve absorption. New oral drugs Vibtex, a biotechnology company co-founded in 2018 by former MIT postdoctoral fellow Thomas von Erlach, MIT Institute Professor Robert Langer, and Traverso, to develop the delivery system, is currently We are pursuing that kind of drug conditioning,” MIT News said.

The researchers said that “the model achieved 100% agreement” in a study using 24 drugs with “well-characterized drug transporter interactions.”

“For 28 clinical drugs and 22 investigational drugs, this model identified 58 unknown drug-transporter interactions, seven of which (out of eight tested) in mice. “This was consistent with kinetic measurements,” they continued.

“We also validated the model's predictions regarding interactions between doxycycline and four drugs (warfarin, tacrolimus, digoxin, and levetiracetam) through ex vivo perfusion assays and analysis of pharmacological data from patients outside the tissue. Screening drugs for interaction with the intestinal transportome via explants and machine learning could help expedite drug development and evaluation of drug safety.”

Giovanni Traverso, associate professor of mechanical engineering at the Massachusetts Institute of Technology and lead author of the study, said: told MIT News “One of the challenges in modeling absorption is that drugs are affected by a variety of transporters.”

“This research depends on how we can model these interactions, which can make drugs safer and more effective and lead to changes that may have previously been difficult to predict. It could help predict potential toxicity,” said Traverso, who is also a gastroenterologist at Brigham University. Women's hospital.

MIT News has more background About the research:

“Previous research has identified several transporters in the gastrointestinal tract that help drugs pass through the intestinal lining. The three most commonly used, which were the focus of the new study, are BCRP , MRP2, and PgP. For this study, Traverso et al. adapted a tissue model they developed in 2020 to measure the absorption of specific drugs. It is based on the intestinal tissue of pigs that have been exposed to a variety of conditions and can be used to systematically expose the tissue to different drug formulations and measure how well they are absorbed.The role of individual transporters within the tissue To study this, the researchers knocked down the expression of each transporter using short RNA strands called siRNA. They knocked down different combinations of transporters in each section of tissue, We were able to study how porter interacts with different drugs.”

The publication explains that to test their predictions, researchers “examined data from approximately 50 patients who were taking any of these three drugs when they were prescribed doxycycline.” found that “when doxycycline was administered to patients already taking warfarin, warfarin concentrations in the patient's bloodstream rose, but fell again when doxycycline was stopped.” “Model predictions that doxycycline absorption is affected by digoxin, levetiracetam, and tacrolimus were also confirmed,” according to the news release.

“There are several avenues that drugs can travel through the organization, but we don't know which one they will take. We can close individual roads to see if drugs will still get through if we close this road. If the answer is yes, then you are not using that road,'' Traverso told the magazine.

“These are commonly used drugs, and we are the first to predict this interaction using this accelerated in silico and in vitro model,” Traverso continued. “This kind of approach allows us to understand the potential safety implications of administering these drugs together.”