Ligand-X Demo
Demo: Relative Binding Free Energy (RBFE)
Video Walkthrough
Coming Soon: Video walkthrough demonstrating the complete RBFE workflow in Ligand-X.
Data Flow Pipeline
When you submit an RBFE calculation, your data flows through these stages:
1. Ligand Preparation
- Parse SDF files, add hydrogens
- Assign AM1-BCC charges using RDKit
2. Ligand Alignment
- 3D alignment using Kartograf atom mapper
- Compute RMSD to reference structure
3. Network Planning
- Build transformation graph (MST/Radial/Maximal)
- Score edges using LOMAP
4. Optional: Docking Validation
- Validate ligand poses via AutoDock Vina
- User confirms poses before proceeding
5. Alchemical Transformations
- For each edge in network: create hybrid topology (OpenFE)
- Run complex leg (protein + ligand)
- Run solvent leg (ligand in water)
- Compute ΔΔG using MBAR
6. MLE Analysis
- Maximum Likelihood Estimation to convert relative ΔΔG values to absolute rankings
Results
- Network graph with ΔΔG for each transformation
- Ranked ligand list with relative binding affinities
- Uncertainty estimates for each value
- Alignment data and atom mapping visualizations
Tools Used
| Stage | Tool | Purpose |
|---|---|---|
| Ligand Prep | RDKit | Molecule parsing, hydrogen addition |
| Charges | Antechamber (AM1-BCC) | Partial charge assignment |
| Atom Mapping | Kartograf | 3D-aware atom mapping |
| Network | LOMAP | Scoring transformations |
| Simulation | OpenMM | GPU-accelerated MD |
| Free Energy | OpenFE | Alchemical RBFE protocol |
| Analysis | PyMBAR | Optimal free energy estimation |
Example Output
After calculation completes, you’ll see:
- Network Viewer: Interactive graph showing all ligand transformations
- Ranked List: Ligands sorted by predicted binding affinity
- Uncertainty: Error bars showing confidence in each prediction
- Convergence: Plots showing how estimates stabilized over simulation time