Vienna LiverTox Workspace

Vienna LiverTox Workspace predicts interaction profiles of small molecules with liver transporters to evaluate inhibition and transport activities relevant to pharmacokinetics, bile flow physiology, drug-drug interactions, and drug-induced liver injury.

Key Features:

• Machine Learning Models: A set of machine learning classification models trained on comprehensive analyses of publicly available data predicts liver transporter interactions.
• Regulatory-Relevant Transporters: Models cover liver transporters recognized as significant by the FDA, EMA, and the Japanese regulatory agency.
• Dual Activity Prediction: Models predict both inhibition and transport activities of small molecules.
• Validation and Predictive Performance: Models were validated using cross-validation techniques and external test sets, achieving balanced accuracies ranging from 0.64 to 0.88.

Scientific Applications:

• Drug Development: Supports early-stage compound prioritization by predicting transporter-mediated effects on pharmacokinetics and efficacy.
• Toxicology Assessment: Informs evaluation of drug-induced liver injury risk by assessing transporter inhibition and transport profiles.

Methodology:

Models were developed from comprehensive analysis of publicly available data using machine learning classification and validated with cross-validation and external test sets reporting balanced accuracies of 0.64–0.88.