AIDrugApp

AIDrugApp predicts inhibitory activity and estimates pIC-50 values for candidate compounds against SARS-CoV-2 targets to support virtual screening and lead selection.

Key Features:

• AI-Based Inhibitory Activity Prediction: Uses deep learning models to classify compounds as active or inactive inhibitors against replicase polyprotein, 3CLpro (main protease), and human angiotensin-converting enzymes.
• pIC-50 Value Estimation: Provides quantitative predictions of pIC-50 values for candidate compounds.
• Virtual Screening Capabilities: Screens large datasets of chemical structures to identify molecules with predicted bioactivity against SARS-CoV-2 targets.
• Data Visualization and Statistics: Generates statistical outputs and visualizations for analysis of screening results.
• Downstream Molecular Docking Support: Integrates with molecular docking analyses to assess binding affinities of screened compounds.
• ADME Analysis Integration: Supports ADME (Absorption, Distribution, Metabolism, and Excretion) analysis for evaluation of pharmacokinetic properties.
• Scalable Data Processing: Architecture supports large-scale processing of molecular datasets.

Scientific Applications:

• Virtual screening for SARS-CoV-2 inhibitors: Rapidly identifies candidate inhibitors against replicase polyprotein, 3CLpro, and human angiotensin-converting enzymes.
• Lead prioritization for experimental validation: Narrows compound sets to prioritize molecules for further biochemical or cellular testing.
• Molecular docking analysis: Assesses predicted binding modes and affinities of top-ranked compounds with viral proteins.
• ADME evaluation: Evaluates pharmacokinetic properties to refine selection of viable drug candidates.

Methodology:

Deep learning models analyze structural and chemical features of compounds; the system supports large-scale data processing and integrates AI-driven predictions with molecular docking and ADME analyses.