Exploring Adverse Event Associations of Predicted PXR Agonists Using the FAERS Database

Area of Science:

• Pharmacology
• Toxicology
• Computational Biology

Background:

• Pregnane X receptor (PXR) is a nuclear receptor regulating drug metabolism and physiological functions.
• The full impact of PXR activation on adverse events remains incompletely understood.
• PXR's role in drug-induced toxicity requires further investigation.

Purpose of the Study:

• To develop a machine learning model for predicting PXR agonist activity.
• To assess the association between PXR agonist activity and adverse drug events.
• To identify potential cardiovascular risks linked to PXR activation.

Main Methods:

• Developed a machine learning model to predict PXR agonist activity.
• Applied the model to drugs in the US Food and Drug Administration Adverse Event Reporting System (FAERS) database.
• Analyzed predicted PXR agonist-drug interactions against reported adverse events using statistical risk assessment (lnROR, -logp).

Main Results:

• The machine learning model successfully predicted PXR agonist activity.
• Statistically significant risks for multiple cardiac disorders were identified for PXR agonist-active drugs.
• Specific drug-cardiac event associations revealed elevated risk profiles.

Conclusions:

• PXR activation is implicated in cardiovascular adverse effects.
• Machine learning models can aid in predicting drug-induced cardiovascular risks.
• Early identification of PXR-mediated risks can enhance drug safety evaluations.