Predicting gastrointestinal drug effects using contextualized metabolic models

摘要: Gastrointestinal side effects are among the most common classes of adverse reactions associated with orally absorbed drugs. These decrease patient compliance treatment and induce undesirable physiological effects. The prediction drug action on gut wall based in vitro data solely can improve safety marketed drugs first-in-human trials new chemical entities. We used publicly available drug-induced gene expression changes to build drug-specific small intestine epithelial cell metabolic models. combination measured silico predicted rates was as features for a multilabel support vector machine predict occurrence showed that combining local wall-specific metabolism performs better than alone, which indicates role development reactions. Furthermore, we reclassified FDA-labeled respect their genetic profiles show hidden similarities between seemingly different linkage xenobiotics transcriptomic could take pharmacology far beyond usual indication-based classifications.