Overview
Based on Computer-Aided Drug Design (CADD) methods, large-scale virtual screening of million-level compound libraries rapidly identifies potentially active molecules, significantly reducing experimental screening cost and time.
Workflow
- Target preparation: Protein structure acquisition and preprocessing (homology modeling, MD optimization)
- Compound library preparation: Collection, standardization, energy minimization of commercial/custom libraries
- Virtual screening:
- Structure-based docking screening (Glide/AutoDock Vina)
- Pharmacophore-based screening
- Machine learning/deep learning scoring models
- Post-processing & prioritization: MM-GBSA binding free energy estimation, ADMET prediction, clustering analysis
- Experimental validation handoff: Recommended compound list for subsequent experimental testing