At Eurofins Discovery, we understand the pressure to accelerate early-stage drug discovery while minimizing risk and cost. This is why our advanced Virtual Screening solutions are designed to streamline hit identification and improve success rates.
To further de-risk your pipeline, our services are enhanced by proprietary ADMET AI models, including SAFIRE and OPLE, which prioritize compounds with favorable pharmacokinetic and safety profiles. The result: a more efficient, data-driven approach to early discovery that empowers your team to make smarter decisions, faster.
Virtual Screening Cascade
Ligand-Based Virtual Screening
Ligand-based virtual screening (LBVS) identifies candidate molecules based on the known activity profiles of previously characterized ligands – without needing the 3D structure of the target protein. In this approach, a target protein is therefore not directly used to identify potential ligands; instead, the emphasis is on finding compounds that are similar in structure and function to known active compounds.
Key features of our ligand-based approach:
- Uses known active compounds or reference ligands as templates
- Similarity searching using:
- → 2D and 3D molecular descriptors
- → Shape-based overlays and electrostatic field comparison
- Pharmacophore modeling to define essential molecular features required for activity
- Virtual screening of ultra-large libraries to discover analogs with improved potency or novel scaffolds
Structure-Based Virtual Screening
Structure-based virtual screening (SBVS) leverages the 3D structure of the biological target (e.g., a protein receptor or enzyme) to identify potential ligands that fit into the active site of the protein. In this approach, a library of small molecules is screened to identify potential ligands that are predicted to bind to the target protein. The method starts with a 3D model of the target protein or a crystal structure of the protein with a bound ligand.
Core elements of our structure-based solutions:
- Utilization of:
- → High-resolution X-ray crystallography and AI-driven protein structure models (including AlphaFold)
- → Molecular docking simulations to fit compounds into the binding pocket
- Binding site identification and preparation using structure analysis tools
- Docking and scoring of diverse compound libraries using:
- → Grid-based docking
- → Flexible ligand conformational sampling
- → Scoring functions for binding affinity estimation
- Post-docking analysis for interaction profiling and hit ranking
Integrated ADMET Profiling with AI Models: SAFIRE and OPLE
To enhance candidate selection, we apply in silico ADMET prediction models in parallel with our virtual screening efforts. These models help identify potential liabilities early and guide medicinal chemistry decisions.
Core elements of our structure-based solutions:
- Solubility
- Permeability
- Plasma Protein Binding (PPB)
- Metabolic Stability
- Efflux potential
- CYP (whole list)
- hERG inhibition (human ether-à-go-go related gene) properties
We have benchmarked our ADMET modeling performance against industry standards, achieving accuracy comparable to top pharmaceutical companies by combining proprietary data with robust machine learning tools.
Why Partner with Eurofins Discovery?
- Extensive chemical space & trusted technology: Access billions of make-on-demand and curated compounds & proven platforms for robust and reproducible results.
- Broad capabilities to rapidly test your predicted molecule in both biological and ADMET assays.
- Decades of experience in novel small molecule discovery
We combine scientific rigor with flexibility to help you move from idea to insight faster and more confidently – reach out to our experts to see how we can support your program.
