What Is SwissDock?
SwissDock 2 is a free web-based molecular docking service from the Swiss Institute of Bioinformatics (SIB). Unlike command-line tools such as AutoDock Vina, SwissDock runs entirely in your browser — you upload your protein structure, draw or upload your ligand, define the binding site, and the docking calculation runs on the SwissDock servers.
SwissDock uses the EADock DSS algorithm, which combines a genetic algorithm for ligand placement with a scoring function based on molecular mechanics energy terms. The service provides estimated binding affinities alongside interactive 3D visualization of docked poses.
The original SwissDock launched in 2011 and has been cited over 2,000 times. SwissDock 2, released in 2023, added an improved interface, faster computation, and better result visualization. It is free for academic use and requires no registration for basic docking.
How to Submit Docking Jobs
SwissDock follows a guided four-step workflow. Here is how to submit a docking job from start to finish.
Step 1: Upload the Protein Structure
SwissDock accepts PDB files. Download your target protein from the Protein Data Bank and upload it. SwissDock will automatically remove water molecules and add hydrogen atoms. For our worked example, we use the ABL1 kinase domain (PDB: 2HYY).
- Navigate to swissdock.ch and click "Start Docking"
- Upload your PDB file by clicking the protein upload area
- SwissDock will display the structure in the 3D viewer
Step 2: Define the Ligand
You can define your ligand in three ways:
- Draw the ligand: Use the built-in molecular editor (powered by MolView) to draw your molecule from scratch
- Upload an SDF or MOL file: If you have a pre-prepared ligand file, upload it directly
- Enter a SMILES string: Paste a SMILES representation and SwissDock will generate the 3D structure
For imatinib, you can draw it in the editor or paste the SMILES: CC1=C(C=C(C=C1)NC(=O)C2=CC=C(C=C2)CN3CCN(CC3)C)NC4=NC=CC(=N4)C5=CN=CC=C5
Step 3: Define the Binding Site
SwissDock needs to know where on the protein to perform docking. You have two options:
- Blind docking: SwissDock explores the entire protein surface. Useful when the binding site is unknown. Slower but comprehensive.
- Site-directed docking: Specify the binding site coordinates or select a region in the 3D viewer. Faster and more accurate when the binding site is known.
For the ABL1-imatinib example, select site-directed docking and center on the ATP-binding pocket (approximately x=28, y=12, z=18 in PDB coordinates).
Step 4: Launch and Monitor
Click "Submit" and SwissDock will queue your job. Processing typically takes 5–15 minutes depending on server load. You can monitor progress in real time and will receive a unique job ID to retrieve results later.
When we submit imatinib against ABL1 kinase (PDB: 2HYY) using site-directed docking, SwissDock identifies the ATP-binding pocket as the optimal binding site. The top-ranked pose shows imatinib nestled in the cleft between the N-lobe and C-lobe, with a predicted binding affinity of approximately −9.2 kcal/mol. The key hydrogen bond to hinge residue Met318 is reproduced in the top 3 poses.
Interpreting Results
Once docking completes, SwissDock presents results in an interactive 3D viewer with multiple poses ranked by predicted binding affinity.
The Results Page
SwissDock displays:
- Ranked poses: Multiple binding orientations sorted by predicted affinity (kcal/mol)
- 3D visualization: Interactive viewer showing protein-ligand interactions, hydrogen bonds, and hydrophobic contacts
- Score breakdown: Estimated ΔG (binding free energy) and predicted Kd (dissociation constant)
- Downloadable files: PDB files of each pose for further analysis in PyMOL or Chimera
What the Scores Mean
| Predicted ΔG | Interpretation | Predicted Kd |
|---|---|---|
| Below −10.0 kcal/mol | Very strong binding | < 1 μM |
| −8.5 to −10.0 kcal/mol | Strong binding | 1–50 μM |
| −6.5 to −8.5 kcal/mol | Moderate binding | 50–500 μM |
| Above −6.5 kcal/mol | Weak or no binding | > 500 μM |
Common Pitfalls
- Don't trust the Kd values blindly: SwissDock's Kd estimates are rough approximations. The correlation with experimental values is moderate at best.
- Focus on pose quality: A pose with clear hydrogen bonds to known catalytic residues is more trustworthy than a high score with no clear interactions.
- Check multiple poses: If the top 3–5 poses show similar orientations (low RMSD), the prediction is more confident.
Compare with Desktop Docking
Run the same imatinib-ABL1 docking using AutoDock Vina and compare the results with SwissDock.
Read the Comparison →Limitations and Considerations
SwissDock is an excellent tool for quick docking studies, but it has important limitations compared to command-line tools like AutoDock Vina or VigyanLLM Docking.
| Feature | SwissDock | AutoDock Vina | VigyanLLM Docking |
|---|---|---|---|
| Installation required | No | Yes | No |
| Flexible receptor | No | Yes | No |
| Batch screening | No | Yes (scripted) | Individual ligands |
| Custom scoring | No | Limited | No |
| Wait times | 5–30 min | 5–10 sec | Instant |
| Internet required | Yes | No | Yes |
| Commercial use | License required | Free (Apache 2.0) | Free |
Key limitations include:
- No flexible receptor docking: SwissDock treats the protein as rigid. For induced-fit effects, use Vina with side-chain flexibility.
- No batch processing: You can dock one ligand at a time. For screening hundreds of compounds, Vina with a shell script is far more efficient.
- Queue dependency: During peak hours, you may wait 30+ minutes for results. Desktop tools like Vina complete in seconds.
SwissDock vs Desktop Tools
SwissDock excels at accessibility — no installation, no command line, and a visual interface that guides you through each step. For students, educators, and researchers who need a quick binding estimate, SwissDock is the fastest path from question to answer.
However, for any serious drug discovery campaign or virtual screening study, a desktop tool like AutoDock Vina or a browser-based tool like VigyanLLM Docking offers significant advantages: faster turnaround, flexible receptor support, and no queue wait times.
For individual ligand studies, we recommend starting with SwissDock to generate initial hypotheses, then validating with Vina or VigyanLLM for more rigorous analysis.
Try VigyanLLM Docking
Same browser-based convenience as SwissDock, but with instant results and no queues.
Open VigyanLLM Docking →Frequently Asked Questions
What is SwissDock?
SwissDock is a free web-based molecular docking service developed by the Swiss Institute of Bioinformatics. It allows users to predict how small molecules bind to protein targets using an intuitive browser interface with no software installation required. The service uses the EADock DSS algorithm and provides estimated binding affinities alongside interactive 3D visualization.
Is SwissDock free?
Yes, SwissDock is free for academic and non-commercial use. Commercial users require a license. The service runs entirely in the browser with no account creation needed for basic docking. For faster results with no queues, VigyanLLM Docking is also free and runs directly in your browser.
How long does SwissDock take?
SwissDock typically completes a single-ligand docking in 5 to 15 minutes depending on server load. During peak hours, queue times may extend to 30 minutes or more. Results are stored for 7 days and can be retrieved via a unique job ID. Command-line tools like AutoDock Vina complete in 5 to 10 seconds.
What are the limitations of SwissDock?
SwissDock has limited parameter control compared to command-line tools like AutoDock Vina. It does not support flexible receptor docking, large-scale virtual screening, or custom scoring functions. Queue-based processing means wait times during peak hours. For batch screening or flexible docking, desktop tools are more appropriate.
How do I interpret SwissDock results?
SwissDock displays docked poses in an interactive 3D viewer with estimated binding affinities in kcal/mol. More negative values indicate stronger predicted binding. Below −10.0 is very strong, −8.5 to −10.0 is strong, −6.5 to −8.5 is moderate, and above −6.5 is weak. Focus on poses with clear hydrogen bonds and good shape complementarity rather than relying solely on the score.
References
- Grosdidier A., Zoete V., Michielin O. (2011). SwissDock, a web-based service to perform molecular docking. Nucleic Acids Research, 39(suppl_2), W248-W254.
- Grosdidier A., Zoete V., Michielin O. (2011). SwissDock: Molecular docking on the web. Swiss Institute of Bioinformatics.
- Trott O., Olson A.J. (2010). AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function. Journal of Computational Chemistry, 31(2), 455-461.