Skip to content
Docking_workshop
Public template

About

No description, website, or topics provided.

Resources

Readme
Activity

Stars

3 stars

Watchers

0 watching

Forks

1 fork
Report repository
BranchesTags

Repository files navigation

nyone can dock, everyone can do!

One molecular docking workflow, four ways to run.

Streamlit App Python 3.10+ License: MIT

Anyone Can Dock makes protein-ligand docking accessible through a zero-install web app, a private local interface, an automation-friendly CLI, and a GPU-ready Google Colab notebook.

From structure preparation to validated docking results, without making molecular docking harder than it needs to be.

🚀 Four ways to use Anyone Can Dock (ACD)

Feature 🌐 ACD Online 🖥️ ACD Local ⌨️ ACD CLI 📓 Google Colab
Best for Quick docking and teaching Private, interactive research Automation and large workflows Workshops and GPU docking
Setup Open a browser Install on your computer pip install anyonecandock Open the notebook
Single and batch docking ✅ ✅ ✅ ✅
Co-Dock (ligands together) ✅ ✅ ✅ ✅
Co-crystal redocking ✅ ✅ ✅ ✅
Receptor-ensemble docking — — ✅ —
Docking engines AutoDock Vina AutoDock Vina AutoDock Vina Vina, VinaXB, GNINA, Vina-GPU
Results and interaction views Interactive 3D and 2D Interactive 3D and 2D Exported structures, scores, and diagrams Notebook 3D views and ProLIF analysis
Automatic Gemini interpretation — — — Optional
Where calculations run Hosted server Your computer Your computer or server Colab runtime
Start here Open Online → Local installation pip install anyonecandock Open in Colab →

Tip: Choose Online for the fastest start, Local for an interactive workflow on your computer, CLI for scripts, or Colab for its additional docking engines. Vina-GPU requires a compatible Colab GPU runtime.

🌐 ACD Online

Best for: beginners, teaching, demonstrations, and quick docking without installation.

  • Runs directly in a web browser
  • Interactive Basic Dock and Batch Dock modes
  • Download structures from RCSB or upload PDB/mmCIF files
  • Automatic, manual, residue-selection, and blind-docking box placement
  • Meeko receptor preparation with Open Babel fallback
  • Metal, heme, cofactor, water, and HETATM handling
  • SMILES, PubChem name, SDF, MOL2, and PDB ligand input
  • pH-aware ligand protonation with pKaNET
  • AutoDock Vina docking and reproducible random seeds
  • Co-crystal redocking and RMSD validation
  • Interactive 3D receptor, binding-box, ligand, and pose visualization
  • Pose Browser with individual pose export
  • Download SDF or PDB poses with explicit hydrogens
  • Preserve all docked poses in the original PDBQT output
  • Interaction diagrams using ACD, RDKit, ProLIF, and PoseView
  • Batch ranking, score tables, ProLIF interaction barcodes, and ZIP export
  • RDKit physicochemical descriptors, drug-likeness rules, and structural alerts
  • Optional ADMET-AI predictions

Streamlit App

🖥️ ACD Local

Best for: private research, full graphical functionality, and unrestricted local workflows.

Includes all major Online features, plus:

  • Runs entirely on the user's own computer
  • Local control of receptor, ligand, result, and temporary files
  • Basic Dock and Batch Dock interfaces
  • Validated ferric-heme and Compound-I preparation
  • Automatic HEME_FERRIC versus CPD_I geometry detection
  • Geometric ferryl oxygen and proximal cysteine identification
  • Fe-O and Fe-S validation gates
  • Ferryl oxygen hydrogen removal
  • JSON-derived RESP charges for Fe, OXO, heme, and proximal cysteine
  • Final PDBQT reread and charge validation before docking
  • Fail-closed heme validation: docking starts only after required checks pass
  • Receptor PDB, scores, individual poses, hydrogenated poses, diagrams, and complete result archives
  • Optional local ADMET-AI analysis
  • No hosted-server runtime or storage restrictions

Linux (Ubuntu/Debian)

sudo apt update
sudo apt install -y \
  python3.11 \
  python3.11-venv \
  openbabel \
  libcairo2-dev \
  libpango1.0-dev \
  libpangocairo-1.0-0

git clone https://github.com/nyelidl/anyone-docking-local.git
cd anyone-docking-local

python3.11 -m venv venv
source venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -r requirements.txt

streamlit run app.py

macOS

brew update
brew install python@3.11 open-babel cairo pango

git clone https://github.com/nyelidl/anyone-docking-local.git
cd anyone-docking-local

python3.11 -m venv venv
source venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -r requirements.txt

streamlit run app.py

Apple Silicon (M1-M4): supported with the native aarch64 Vina binary.

Windows

Recommended: use WSL2 with Ubuntu and follow the Linux instructions above.

For native Windows:

  1. Install Open Babel and add it to PATH.
  2. Install Cairo/Pango, for example with conda install -c conda-forge cairo pango.
  3. Clone and run the app:
git clone https://github.com/nyelidl/anyone-docking-local.git
cd anyone-docking-local
python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt
streamlit run app.py

⌨️ ACD CLI

Best for: automation, reproducible research, scripting, servers, and large screening projects.

  • Commands for docking, ligand batch docking, receptor-ensemble docking, receptor preparation, ligand preparation, redocking, and diagrams
  • Accepts PDB IDs or local PDB/mmCIF structures
  • Accepts SMILES, PubChem names, SDF, MOL2, and PDB ligands
  • Automatic co-crystal, manual, selection-based, and blind-docking boxes
  • No artificial docking-box size limit
  • Preparation-only mode without starting Vina
  • Prepared-receptor JSON export and reuse
  • pH-aware, tautomer-aware ligand preparation using pKaNET
  • Microstate selection and configurable tautomer limits
  • Deterministic conformer and Vina random seeds
  • AutoDock Vina exhaustiveness, pose count, and energy-range controls
  • Batch-safe noninteractive operation
  • Protein-ensemble docking from receptor files, directories, glob patterns, or PDB IDs
  • Independent preparation, box generation, and heme/Compound-I validation for every ensemble member
  • Per-receptor failure isolation with combined raw-score and statistical summary CSV files
  • Automatic co-crystal redocking and heavy-atom RMSD verdicts
  • Validated ferric-heme and Compound-I preparation
  • RESP charge assignment and final PDBQT validation
  • Multi-heme-center support
  • PDBQT, SDF, bond-order-corrected SDF, PDB, CSV, and SVG output
  • Eight interaction classes, including hydrogen bonding, hydrophobic, aromatic, ionic, halogen, and metal/heme coordination
  • Python API access for integration into custom pipelines

Install from PyPI

pip install anyonecandock

Commands

acd <command> [options]

  dock      Full pipeline: receptor + ligand + Vina (single ligand)
  redock    Self-docking validation; no SMILES input needed
  batch     Dock multiple ligands against one receptor
  ensemble  Dock multiple ligands against multiple receptor conformations
  receptor  Prepare receptor only
  ligand    Prepare ligand only
  diagram   Generate a 2D interaction diagram SVG

Quick examples

# Dock a single ligand by compound name
acd dock --receptor 1M17 --compound erlotinib

# Dock by SMILES
acd dock --receptor 4AGN --smiles "CCO" --name ethanol

# Reuse a prepared receptor
acd dock --receptor-json ./rec/receptor_summary.json --compound baicalein

# Dock from local structure files
acd dock --receptor structure.cif --ligand-file ligand.sdf --name mylig

# Self-docking validation
acd redock --receptor 4AGN
acd redock --receptor structure.cif --fmt CIF
acd redock --receptor 4AGN --resname DC3 --diagram

# Batch docking
acd batch --receptor 1M17 --ligands compounds.smi
acd batch --receptor 4AGN --smiles-list "CCO ethanol" "c1ccccc1O phenol"

# Receptor-ensemble docking
acd ensemble --receptors "ensemble/*.pdb" --ligands compounds.smi --seed 42

# Ensemble docking with aligned receptors and a shared box
acd ensemble \
  --receptors model1.pdb model2.cif model3.pdb \
  --ligands compounds.smi \
  --center manual --cx 14.2 --cy -3.8 --cz 22.1 \
  --bx 22 --by 22 --bz 22 \
  --seed 42 \
  --output ./ensemble_screen

Receptor ensembles

The ensemble workflow accepts local PDB/mmCIF files, directories, quoted glob patterns, and four-character PDB IDs. It runs the established ACD batch workflow independently for every receptor, so receptor preparation and validation are never shared or bypassed between conformations.

For heme-containing structures, each ensemble member independently undergoes ferric-heme or Compound-I detection, Fe-S/Fe-O geometry validation, RESP assignment, and final PDBQT reread validation. A receptor that fails preparation or geometry validation is recorded and skipped rather than passed to Vina.

Use automatic or selection-based centering when each receptor requires its own box. A shared manual box should be used only when all receptor conformations are structurally aligned in the same coordinate frame.

ensemble_screen/
  001_model1/                 Complete batch output for receptor 1
  002_model2/                 Complete batch output for receptor 2
  ensemble_receptors.csv      Receptor preparation and run status
  ensemble_scores.csv         All receptor-ligand score rows
  ensemble_summary.csv        Best, mean, median, and population SD per ligand

The ensemble statistics are transparent ranking aids. They are not population-weighted thermodynamic binding energies.

Redocking verdict

The redock workflow automatically identifies and re-docks the co-crystal ligand. SMILES are obtained in this order:

  1. RCSB Chemical Component Dictionary (CCD)
  2. CIF _chem_comp.pdbx_smiles
  3. 3D coordinate conversion as a fallback

The heavy-atom RMSD verdict is reported automatically:

  • PASS: RMSD <= 2.0 Å
  • BORDERLINE: RMSD > 2.0 Å and <= 3.0 Å
  • FAIL: RMSD > 3.0 Å

Python API

from anyonecandock import core

SEED = 72
BOX_SIZE = (18, 18, 18)
vina_bin = "/path/to/vina"

result = core.prepare_receptor(
    raw_pdb="raw.pdb",
    wdir="./rec",
    box_size=BOX_SIZE,
)

lig = core.prepare_ligand(
    smiles="c1ccc(cc1)O",
    name="phenol",
    ph=7.4,
    wdir="./lig",
    mode="pkanet",
    use_pubchem=False,
    max_tautomers=8,
    ph_window=1.0,
    conformer_seed=SEED,
)

dock = core.run_vina(
    receptor_pdbqt=result["rec_pdbqt"],
    ligand_pdbqt=lig["pdbqt"],
    config_txt=result["config_txt"],
    vina_path=vina_bin,
    exhaustiveness=16,
    n_modes=10,
    energy_range=3,
    seed=SEED,
    wdir="./out",
    out_name="phenol",
)

print("Top score:", dock["top_score"])

📓 ACD Google Colab

Best for: workshops, education, portable research, and GPU docking without local setup.

  • Guided notebook workflow with step-by-step preparation and analysis
  • Upload PDB/mmCIF structures or download them from RCSB
  • Receptor and pH-aware ligand preparation
  • Validated ferric-heme and Compound-I handling
  • Four supported docking engines:
    • AutoDock Vina 1.2.7
    • VinaXB
    • GNINA
    • Vina-GPU 2.1
  • Google T4 GPU support for Vina-GPU
  • Co-crystal redocking before production docking
  • Interactive ligand and pose selection
  • Binding-box validation
  • Heavy-atom RMSD analysis across poses
  • Engine-aware score extraction
  • Ranked CSV results and score plots
  • Individual result download or complete ZIP packaging
  • Mobile- and tablet-friendly result export
  • No additional prepare_receptor.py upload required; required preparation logic is included in the notebook/package

Open In Colab

📊 Choose Your Version

Capability Online Local CLI Colab
Graphical workflow Yes Yes No Notebook
No local installation Yes No No Yes
Basic docking Yes Yes Yes Yes
Batch docking Yes Yes Yes Yes
AutoDock Vina Yes Yes Yes Yes
VinaXB, GNINA, and Vina-GPU No No No Yes
Automated scripting No No Yes Partial
Interactive Pose Browser Yes Yes File output Yes
ADME/ADMET analysis Yes Yes Optional tools No
Strict ferric/Compound-I validation Yes Yes Yes Yes
Private local processing No Yes Yes Colab runtime

🧬 Heme-aware docking

ACD Local, CLI, and Colab support validated preparation of ferric heme and Compound-I systems.

For supported heme centers, ACD can:

  • Detect HEME_FERRIC versus CPD_I from local Fe coordination geometry
  • Identify a ferryl oxygen geometrically rather than relying only on atom names
  • Recognize ferryl oxygen labels such as O, O1, OXY, or OXO
  • Identify the proximal cysteine sulfur geometrically
  • Validate Fe-O and Fe-S distances before docking
  • Remove an incorrectly retained O-H hydrogen from the ferryl oxygen
  • Apply JSON-derived RESP charges to Fe, OXO, the heme group, and proximal cysteine
  • Re-read the final PDBQT and validate required charges before docking
  • Fail closed if required heme-state checks do not pass
  • Handle multiple heme centers in CLI workflows

🗺️ Interaction analysis

ACD supports eight interaction classes in its geometry-based analysis, including:

  • Hydrogen bonds
  • Hydrophobic contacts
  • Aromatic interactions
  • Cation-pi interactions
  • Ionic interactions
  • Metal/heme coordination
  • Halogen bonds
  • H···halogen contacts

Depending on the interface, results can also be visualized with RDKit, ProLIF, and PoseView.

🧬 Supported protein types

Protein class Support Notes
Standard single-chain proteins ✅ Full Primary use case
Multi-chain / homo-oligomers ✅ Full Multi-chain structures supported
Heme proteins ✅ Full Ferric heme and Compound-I supported in Local/CLI/Colab
Metal-binding proteins ✅ Full Common metal ions can be retained during preparation
MD simulation outputs ✅ Full PDB structures from common MD workflows can be used
Non-standard ligand names ✅ Full Ligand-like HETATM records can be detected
Modified amino acids ✅ Full Supported when retained as part of the receptor
Multiple co-crystal ligands ✅ Full Ligand selection supported
Cofactor-binding proteins ✅ Full Cofactors can be retained or removed as appropriate
RNA / DNA targets ⚠️ Partial No nucleic-acid-specific interaction model
Covalent docking ❌ No Standard Vina docking is non-covalent

🤖 AI interfaces

AI interfaces are optional front ends to ACD rather than separate docking engines.

Anyone Can Dock GPT

Ask ChatGPT to dock molecules using natural-language instructions.

ChatGPT

How to do: https://scribehow.com/o/an3IFiDaRtmZANATDZDBFg/viewer/How_to_Dock_Gefitinib_to_PDB_1M17__mrZrG5ZmTbSab6g-FxdVRQ

Anyone Can Dock in Claude

Connect Claude to the ACD API as a custom MCP connector.

MCP server URL:

https://anyone-can-dock-mcp.anyonecandock.workers.dev

Add to Claude

Example prompts:

"dock quercetin into JAK2 and report binding affinity"
"compare erlotinib vs gefitinib binding to EGFR (1M17)"

💻 Platform compatibility

Platform Vina binary Open Babel Status
Linux x86_64 Auto-download apt install openbabel Fully supported
macOS Intel Auto-download brew install open-babel Fully supported
macOS Apple Silicon Native aarch64 brew install open-babel Fully supported
Windows x86_64 Auto-download Installer / WSL2 Supported; WSL2 recommended
Streamlit Cloud Auto-download via packages.txt Fully supported
Google Colab Auto-download apt install openbabel Fully supported

📄 Citation

If you use ACD in research, please cite the relevant methods used in your workflow.

Anyone Can Dock: An Online Molecular Docking Tool for Everyone
Hengphasatporn, K.; Bunchuay T.; Duan, L.; Shigeta, Y., J. Cheminformatics., 2026. DOI: 10.1186/s13321-026-01306-2

AutoDock Vina 1.2.7
Eberhardt et al., J. Chem. Inf. Model., 2021. DOI: 10.1021/acs.jcim.1c00203

DFDD
Hengphasatporn, K.; Duan, L.; Harada, R.; Shigeta, Y., J. Chem. Inf. Model., 2026. DOI: 10.1021/acs.jcim.5c02852

RDKit · Landrum, G. (2023) · https://www.rdkit.org

ProDy · Bakan et al., Bioinformatics, 2011. DOI: 10.1093/bioinformatics/btr168

stmol · Nápoles-Duarte et al., Front. Mol. Biosci., 2022. DOI: 10.3389/fmolb.2022.990846

Dimorphite-DL · Ropp et al., J. Cheminform., 2019. DOI: 10.1186/s13321-019-0336-9

pKaNET Cloud · Please cite the corresponding manuscript when available.

gemmi (optional, for CIF support) · Wojdyr, M., JOSS, 2022. DOI: 10.21105/joss.04200

📜 License

MIT License — see LICENSE for details.

Anyone Can Dock: from structure preparation to validated docking results, without making molecular docking harder than it needs to be.

About

No description, website, or topics provided.

Resources

Readme
Activity

Stars

3 stars

Watchers

0 watching

Forks

1 fork
Report repository

Releases

Packages

Contributors

Languages