ToxSTAR

ToxSTAR predicts human drug- and chemical-induced toxicity, with a focus on drug-induced liver injury (DILI), by leveraging structural similarity analysis and computational models grounded in biotechnology (BT) and information technology (IT).

Key Features:

• Structural similarity-based analysis: Performs structural similarity assessments of drugs and their metabolites, including consideration of reactive metabolites, to identify structural configurations associated with toxicity.
• Metabolite-aware molecular evaluation: Examines molecular structures of parent compounds and metabolites to assess potential contributions to hepatotoxicity.
• In-house DILI subtype prediction models: Provides predictive models that classify four DILI subtypes: cholestasis, cirrhosis, hepatitis, and steatosis.
• Integration of structural and computational modeling: Combines structural similarity assessments with advanced computational modeling to enhance the precision and reliability of toxicity predictions.
• Focus on human toxicity endpoints: Targets prediction of human-relevant toxicity outcomes for drugs and chemicals, addressing multifactorial mechanisms of liver damage.

Scientific Applications:

• Drug development and safety assessment: Enables evaluation of hepatotoxic potential of new chemical entities to inform safety decision-making in drug development.
• Mechanistic DILI research: Supports investigation of mechanisms underlying DILI, including the roles of reactive metabolites and diverse pathways leading to cholestasis, cirrhosis, hepatitis, and steatosis.

Methodology:

Computational methods explicitly include structural similarity assessments of molecular structures and metabolites and in-house predictive modeling for the four DILI subtypes (cholestasis, cirrhosis, hepatitis, steatosis).