What is AutoDock Vina?
AutoDock Vina is a molecular docking program that predicts how a small molecule binds to a protein by searching through possible binding poses and scoring them with an empirical force field. It is the most widely used open-source docking tool, combining fast search with practical accuracy for routine structure-based drug discovery.
On ProteinIQ, the Vina interface exposes the full Vina family workflow rather than just the original 2010 defaults. This includes the standard Vina scoring function, the Vinardo variant optimized for virtual screening, and an AutoDock4 scoring option for metal coordination or legacy workflows.
Applications
- Predicting ligand binding modes in a known binding pocket
- Ranking a focused compound set against one receptor (virtual screening)
- Comparing binding hypotheses after small structural changes to prioritize compounds
- Testing whether a proposed binding site can accommodate a ligand with plausible geometry
How to use AutoDock Vina online
ProteinIQ runs AutoDock Vina in the browser with cloud execution, so receptor preparation, ligand submission, and docking outputs are available without local installation or command-line setup.
Inputs
| Input | Description |
|---|
Protein (Receptor) | Upload a receptor structure as .pdb or .ent, or fetch one by 4-character PDB ID such as 1HSG. Protein atoms are required, and the validator warns when the uploaded file appears to contain nucleic acids or YASARA-specific formatting. |
Ligand | Provide ligand input as SMILES text, a supported structure file (.pdbqt, .sdf, .mol, .mol2, .smiles, .smi, .txt, .csv), or fetch by PubChem. One ligand is used in Single ligand mode, up to 5 ligands in Simultaneous co-docking, and up to 10 entries in Batch. |
Vina docking mode | Single ligand runs one docking job, Simultaneous co-docking places multiple ligands in the same search space during one run, and Batch docks each ligand independently against the same receptor. |
Job name | Optional label stored with the run to make repeated docking experiments easier to identify. |
Ligand validation blocks metal-containing ligands and disconnected multi-fragment submissions in the standard Vina workflow. Those cases are better suited to GNINA or to a more specialized docking setup.
Settings
| Setting | Description |
|---|
Scoring function | Vina is the default general-purpose choice. Vinardo uses a modified empirical model often favored for virtual screening benchmarks. AutoDock4 uses the classical AutoDock4 force-field-style scoring and is the required option for hydrated ligand and zinc-specific workflows in this interface. |
Docking mode | Dock performs a full search. Score only evaluates the submitted pose without searching. Local only refines the input pose locally. Randomize only generates randomized ligand placement without full docking output ranking. |
Exhaustiveness | Search thoroughness from 1 to 64, with a default of 8. Larger values increase runtime but improve the chance of recovering lower-energy poses. |
Number of poses | Maximum number of docked poses returned, from 1 to 50. |
Results
| Output | Description |
|---|
Viewer | Interactive 3D view of receptor and docked pose geometry |
Data | Table of scored poses with binding affinity (kcal/mol) and links to output files |
Files | Downloadable structure files for the generated docking poses |
In batch runs, each ligand is processed separately under the same receptor and parameter set, which makes the output suitable for side-by-side ranking.
How does AutoDock Vina work?
AutoDock Vina combines an empirical scoring function with stochastic global search and gradient-based local optimization. The ligand is translated, rotated, and flexed inside a predefined search volume; each candidate pose is scored; and promising conformations are refined before final clustering and ranking.
Scoring
The scoring function estimates binding favorability using weighted steric, hydrophobic, hydrogen-bonding, and penalty terms. Vina uses the default empirical model. Vinardo modifies the parameterization for improved enrichment in some screening benchmarks. AutoDock4 uses older force-field-style terms and is required when metal-aware behavior or hydrated ligand workflows are needed.
Vina-family affinity scores are most useful as relative rankings within a single consistent experiment. Absolute kcal/mol values should not be treated as direct binding free energy estimates, especially across different targets, protonation states, or receptor preparations.
Search
Vina uses multiple independent runs initialized with random poses. Within each run it alternates broad perturbation steps with local refinement, then collects the best solutions across all searches. The Exhaustiveness parameter controls total independent search effort, while Number of poses and Energy range determine how many alternatives survive to the final report.
Flexible residue docking
Most docking calculations keep the receptor rigid apart from ligand torsions. When Flexible residues are specified, selected receptor side chains are allowed to move during the search. This can recover poses that rigid docking would miss, but increases the search space and runtime substantially. Flexible docking is usually reserved for a few residues with a clear mechanistic rationale for moving.
Interpreting results
The affinity score is most useful as a relative ranking within one consistent experiment. Absolute kcal/mol values should not be treated as direct binding free energies, especially across different targets, protonation states, or receptor preparations.
| Affinity range | Typical interpretation |
|---|
| < -10 kcal/mol | Very favorable predicted binding |
| -7 to -10 kcal/mol | Strong docking score worth follow-up |
| -5 to -7 kcal/mol | Moderate score that often needs visual inspection |
| > -5 kcal/mol | Weak or uncertain binding hypothesis |
Pose geometry matters as much as score. A slightly worse score with sensible hydrogen bonding, steric fit, and ligand burial is often more credible than the top-ranked pose if that pose shows clashes or unrealistic exposure. For batch docking, comparisons are most meaningful when all ligands were prepared with the same protonation and tautomer assumptions.
Limitations
- Rigid receptor approximation: Whole-protein backbone motion is not modeled, so induced-fit effects can still be missed even when flexible residues are enabled.
- Approximate scoring: Docking scores support prioritization, not definitive affinity prediction.
- Preparation sensitivity: Protonation state, tautomer choice, bound waters, and search-box placement can change rankings substantially.
- Metal handling: Standard Vina scoring is not a good default for metal-dependent binding chemistry—use GNINA or AutoDock4 for those cases.
- Large or highly flexible ligands: Search quality drops as conformational complexity increases, even when runtime is increased.
- Fragment handling: Standard Vina blocks metal-containing ligands and disconnected multi-fragment submissions—use GNINA for those cases.
- GNINA: Vina-derived docking with convolutional neural network rescoring and better support for metal-containing ligands
- DiffDock: Diffusion-based docking that can be useful when the binding site is uncertain
- AutoDock-GPU: High-throughput AutoDock workflow for larger screening workloads
- HADDOCK3: Integrative docking platform for complexes that benefit from experimental restraints
