A Robust, Mechanistically Based In Silico Structural Profiler for Hepatic Cholestasis

Area of Science:

• Computational Toxicology
• Medicinal Chemistry
• Drug Metabolism

Background:

• The liver's central role in xenobiotic metabolism makes it susceptible to drug-induced injury.
• Cholestasis, a key indicator of liver dysfunction, poses significant challenges in pharmaceutical development.
• Existing in vitro methods for predicting cholestasis have limitations in accuracy and reliability.

Purpose of the Study:

• To develop a computational toxicology tool (in silico profiler) for predicting drug-induced cholestasis.
• To identify specific structural alerts associated with cholestasis using clinical drug data.
• To enhance early-stage drug safety assessment in pharmaceutical research.

Main Methods:

• Utilized a dataset of over 1500 small molecular drugs.
• Formulated 15 distinct structural alerts based on clinical data and known drug classes.
• Analyzed structural motifs, reactive mechanisms, and structure-activity relationships.

Main Results:

• Identified 15 structural alerts reliably associated with cholestasis.
• These alerts cover diverse pharmaceutical classes, including psychoactive tricyclics, beta-lactam antimicrobials, and steroids.
• Provided detailed descriptions of alert coverage, selectivity, and underlying chemical mechanisms.

Conclusions:

• The developed in silico profiler effectively identifies structural motifs linked to drug-induced cholestasis.
• This computational approach offers a valuable tool for early risk assessment in drug discovery.
• Mechanistic insights provided by the alerts facilitate understanding within the adverse outcome pathway framework.