Before orally administered drugs can make their way throughout the body, they must first bind to membrane proteins called drug transporters, which carry compounds across the intestinal tract and help them reach their intended targets. But because one drug can bind to several different drug transporters, they may struggle to get past this gut barrier, potentially leading to decreased drug absorption and efficacy. If another drug is added to the mix, interactions between the two compounds and their transporters can cause dangerous side effects.

Researchers from Brigham and Women’s Hospital, a founding member of the Mass General Brigham healthcare system, and MIT have designed a model that analyzes the flow of drugs through tissues and uses machine learning to predict how specific compounds will interact with different transporters. When they used pig tissue to test their machine learning model on 50 approved and investigational drugs, they identified 58 previously unknown drug-transporter interactions and 1,810,270 unknown potential interactions between different drugs.

"Our model has the potential to help accelerate drug discovery and allow drugmakers to better understand safety concerns associated with mixing different medicines," said senior author Giovanni Traverso, MD, PhD, MBBCH a gastroenterologist in the Brigham's Division of Gastroenterology, Hepatology, and Endoscopy.

Shi Y, Reker D, Byrne JD, Kirtane AR, Hess K, Wang Z, Navamajiti N, Young CC, Fralish Z, Zhang Z, Lopes A, Soares V, Wainer J, von Erlach T, Miao L, Langer R, Traverso G.
Screening oral drugs for their interactions with the intestinal transportome via porcine tissue explants and machine learning.
Nat Biomed Eng. 2024 Feb 20. doi: 10.1038/s41551-023-01128-9