New Test Predicts Liver Drug Problems Better Than Old Methods

Drug companies need ways to catch liver problems before medicines reach patients. One serious issue involves something called the bile salt export pump or BSEP. When drugs block this pump, it can lead to dangerous liver buildup of bile acids. Scientists wanted to find better prediction methods for this problem. They studied 62 different compounds that already had known effects on BSEP. The team built mathematical models using multiple linear regression to see what works best. Traditional approaches used simple measures like how lipophilic a molecule is or its molecular weight. These older methods don't really show what happens at the cell membrane where BSEP lives. The researchers tested biomimetic chromatographic descriptors instead. These come from special lab techniques that mimic biological membranes and measure how strongly drugs bind to human serum albumin. T

his approach captures more realistic conditions inside the body. The results showed biomimetic models outperformed traditional ones significantly. They were more balanced, more robust, and better at predicting new compounds. The combination of membrane affinity and protein binding data gave a clearer picture of why some drugs interact with BSEP. This matters because it represents what actually happens near the canalicular membrane where the pump operates. Conventional descriptors miss this crucial location. The new method goes beyond standard chemical structure-activity relationships by including distribution properties that affect drug behavior. It provides a straightforward, interpretable system for identifying potential transporter-mediated toxicity early in drug development. This approach could save time and prevent costly failures later in the research process.

actions