High-Throughput Virtual Screening & In Silico Drug Design
Triage massive chemical libraries of 500,000+ compounds down to nanomolar-affinity leads. Our structure-based drug discovery workflow combines physics-based docking, machine-learning QSAR, and rigorous MM-GBSA free energy rescoring.
Input Requirements & Target Preparations
Our computational chemists work with experimental and computational target structures:
Macromolecular Targets
- High-resolution X-ray crystallography (PDB) and Cryo-EM models
- AlphaFold2 and AlphaFold3 predicted apo and holo conformations
- Cofactor and metal coordination assignment with proper protonation states
Chemical Libraries
- SDF, SMILES, and MOL2 formats (ZINC20, Enamine, Mcule, ChemSpace)
- Proprietary client internal compound libraries under bilateral NDA
- In silico combinatorial library enumeration and tautomer generation
Multi-Stage Virtual Screening Pipeline
We execute a hierarchical filtration strategy that balances computational efficiency with biophysical precision:
- Physicochemical Pre-Filtering: Lipinski Rule of 5, Veber oral bioavailability rules, and PAINS (Pan-Assay Interference Compounds) elimination via RDKit algorithms.
- Ensemble Molecular Docking: Multi-conformation docking across flexible receptor states using AutoDock Vina and Glide SP/XP, accounting for induced-fit binding pocket plasticity.
- Post-Docking MM-GBSA Rescoring: Molecular Mechanics Generalized Born Surface Area (MM-GBSA) calculations to estimate true thermodynamic binding free energies (ΔGbind) and discard false-positive docking hits.
- Predictive QSAR & ADMET Profiling: Random forest and graph neural network models predicting human ether-a-go-go (hERG) cardiac toxicity, blood-brain barrier permeability, CYP450 metabolic inhibition, and plasma protein binding.
Turnaround & Deliverables Package
Within 3 to 4 weeks, our computational team delivers actionable medicinal chemistry portfolios:
- Rank-ordered candidate compound spreadsheet with predicted affinities (kcal/mol), chemical identifiers, and SMILES.
- 3D PyMOL (.pse) and Mol* interactive session files displaying key hydrogen bonds, salt bridges, and pi-stacking interactions.
- ADMET radar plots and pharmacophore alignment maps.
- Purchasing manifest with Enamine/ZINC catalog numbers for immediate wet-lab assay ordering.
Launch Your Virtual Screening Campaign
Consult with our computational chemistry team under a mutual non-disclosure agreement to establish docking grids and compound screening thresholds.
Submit Technical Scoping Inquiry →