MCT8 Docking Documentation
Getting Started
What is this tool? This web application predicts the binding affinity between small molecules and MCT8 (Monocarboxylate Transporter 8), a critical protein in thyroid hormone transport.
Your First Docking Simulation
- Enter a molecule: Use the default SMILES string or paste your own
- Set parameters: Start with defaults (3 poses, exhaustiveness 8)
- Run docking: Click "Run Docking" and wait 1-5 minutes
- Review results: Check the binding affinity score and assessment
- Visualize in 3D: Click any result row to see the binding pose
Quick Tips: Lower (more negative) binding affinity = stronger binding | Scores below -9.0 kcal/mol indicate likely MCT8 inhibition | Use multi-pose view to compare top 5 binding modes
Input Guide
SMILES Strings
SMILES (Simplified Molecular Input Line Entry System) is a text notation for molecules.
Examples: CCO (Ethanol) | c1ccccc1 (Benzene)
File Upload Formats
- SDF (.sdf): Pre-prepared 3D structures (recommended for multiple molecules)
- SMILES (.smi, .txt): One SMILES per line
- CSV (.csv): Must contain a 'SMILES' column
Ligand Preparation Options
Add Hydrogens: Essential for accurate docking (keep checked) | Generate 3D: Creates 3D coordinates from SMILES (recommended) | Optimize Geometry: MMFF force field optimization (recommended)
Docking Parameters Explained
Number of Poses (1-50, default: 3)
How many different binding orientations to generate. More poses = better sampling but longer computation time.
Exhaustiveness (2-128, default: 8)
Search thoroughness. Higher values increase accuracy but take longer. Use 8 for standard runs, 16+ for publication-quality results.
Search Box (0-10 Å, default: 6)
Expansion around the binding site. Larger values allow more exploration but increase search space. 6Å is optimal for most cases.
CNN Scoring (none/fast/default, default: none)
Neural network rescoring. Set to "none" for CPU-only mode. Use "fast" or "default" if you have a GPU for improved accuracy.
Results Interpretation
Binding Affinity Scale
- < -9.0 kcal/mol: Likely Inhibitor (High developmental risk)
- -8.0 to -9.0 kcal/mol: Possible Inhibitor (Moderate risk)
- > -8.0 kcal/mol: Unlikely Inhibitor (Low risk)
Understanding Scores
Binding Affinity: Estimated free energy of binding (kcal/mol). More negative = stronger binding = higher likelihood of MCT8 inhibition.
CNN Score: Neural network confidence score (0-1) when CNN scoring is enabled. Higher = more confident prediction.
Boltzmann Weight: Thermodynamic probability of each binding pose. Higher weights indicate more energetically favorable conformations.
3D Viewer Help
Opening the Viewer
Click any row in the results table to open the 3D molecular viewer showing the receptor, binding site, and ligand pose.
Mouse Controls
- Rotate: Left-click and drag
- Zoom: Scroll wheel or pinch gesture
- Pan: Right-click and drag
Display Options
Show Receptor Cartoon: Toggle protein backbone visualization | Show Binding Site: Highlight key residues in the binding pocket | Multi-Pose View: Overlay top 5 poses in different colors for comparison
API Documentation
Endpoint: POST /api
Request:
{
"smiles": ["CCO", "c1ccccc1"],
"params": {
"num_modes": 3,
"exhaustiveness": 8,
"autobox_add": 6.0,
"cnn": "none"
},
"format": "json"
}Response:
{
"success": true,
"num_poses": 6,
"results": [
{
"pose_index": 0,
"smiles": "CCO",
"affinity": -5.2,
"assessment": "Unlikely Inhibitor"
}
]
}Output Formats
"format": "json" - JSON response (default) | "format": "csv" - CSV file | "format": "sdf" - SDF file with 3D structures
Additional Endpoints
GET /api/receptor - MCT8 protein structure | GET /api/binding_site - Binding site coordinates | GET /api/pose/{pose_id} - Individual pose data
Troubleshooting
Invalid SMILES Error
Solution: Verify SMILES syntax using a chemical structure editor like PubChem or ChemDraw. Ensure proper capitalization (C vs c for aromatic).
Docking Timeout
Solution: Reduce exhaustiveness parameter or number of poses. Large molecules may require longer computation time.
3D Viewer Not Loading
Solution: Ensure WebGL is enabled in your browser. Try a different browser (Chrome/Firefox recommended). Clear browser cache.
Multi-Pose Colors Not Showing
Solution: Ensure you've run docking first. Toggle the multi-pose checkbox. If issues persist, refresh the page and try again.
CSV Upload Format Error
Solution: Ensure your CSV has a column named exactly "SMILES" (case-sensitive). One SMILES string per row.
Docker Container Won't Start
Solution: Check if port 5000 is already in use: lsof -i :5000. View logs: docker logs mct8-docking-app
About This Tool
Project Information
The MCT8 Binding Affinity Prediction tool is part of the Virtual Human Platform for Safety (VHP4Safety) initiative, developing computational models to predict chemical toxicity and reduce animal testing.
Technology Stack
Backend: Python 3.10+, Flask, RDKit | Docking Engine: Gnina v1.3 | Frontend: HTML5, JavaScript, 3Dmol.js | Deployment: Docker (CPU-only)
Citation
[TO-DO]
Contact
Website: www.vhp4safety.nl | GitHub: github.com/VHP4Safety