Drug efficacy in humans can be heavily influenced by how it interacts with various digestive system transporters. Researchers at the Massachusetts Institute of Technology (MIT), Duke University, and Brigham and Women’s Hospital have developed a method that identifies the interactions between drugs and these transporters. These interactions can potentially result in adverse effects if two drugs that use the same transporter pathway are taken in combination.
A multidimensional approach has been used to understand these interactions, employing tissue models and machine learning algorithms. The new research has indicated that an antibiotic and a blood thinner taken together can negatively affect each other’s efficacy, since they use the same transporter protein in the digestive system.
The researchers conducted this study with the intention of predicting drug interactions that could cause possible toxicities, thereby enhancing drug safety and effectiveness. This knowledge could be especially valuable for drug developers, who could use additives to improve a drug’s absorbability by helping it interact optimally with transporter proteins.
The research began with a tissue model originally developed by the scientists in 2020. This pig intestinal tissue model grown in a lab helped measure how drug absorption can be influenced by different drug formulations. To understand the roles of individual transporters within the tissue, the team used short RNA strands (siRNA) to reduce the activity of each transporter. They then analyzed the interactions of each transporter with diverse drugs.
They subjected 23 widely used drugs to this system, which helped identify the transporters used by each. Using this data, a machine learning model was trained to predict the interactions between drugs and transporters on the basis of the chemical structures of the drugs. This model analyzed a fresh set of 28 drugs in current use and 1,595 experimental drugs, resulting in almost 2 million predictions of potential drug interactions.
The researchers verified their model’s predictions by analyzing data from patients who were prescribed certain drug combinations. For instance, they found that patients who were prescribed the antibiotic doxycycline along with the blood thinner warfarin had temporarily elevated levels of warfarin in their bloodstream—exactly as the model had predicted.
This study marks a significant development in understanding potential drug interactions and verifying this new study model’s effectiveness. Future applications may include pre-approval research on the absorbability of new drugs, identifying probable dangerous interactions, and aiding in drug development to reduce unforeseen toxicities and improve the efficacy of treatment.
