SuperCYPsPred

SuperCYPsPred predicts cytochrome P450 (CYP) enzyme inhibition and specificity from 2D chemical structures to assess drug metabolism and potential drug–drug interactions.

Key Features:

• Focus on Major Isoforms: Targets CYP1A2, CYP2C9, CYP2C19, CYP2D6, and CYP3A4, which together account for over 80% of clinical drug metabolism.
• Machine Learning-Based Predictions: Employs machine learning models validated by cross-validation and external validation sets to predict CYP inhibition profiles.
• Input and Output: Accepts 2D chemical structures and outputs CYP inhibition profiles from ten different models using various molecular fingerprints, confidence scores, lists of similar compounds with known interactions, and detailed per-cytochrome interaction profiles including DDI tables.
• Literature Integration: Provides curated literature details on drug–cytochrome interaction networks.

Scientific Applications:

• DDI identification: Assists identification of potential drug–drug interactions during early-stage drug design.
• ADMET assessment: Provides insights into Absorption, Distribution, Metabolism, Excretion, and Toxicity properties relevant to drug development.
• Regulatory support: Supports regulatory assessments by predicting enzyme interactions that could lead to adverse effects.

Methodology:

Machine learning algorithms trained on datasets of known drug-CYP interactions using molecular fingerprints, implemented as ten different models, with performance assessed by cross-validation and external validation and outputs including confidence scores and similarity-based compound retrieval.