Quick Start

This tutorial will get you up and running with PandaDock in minutes.

Installation

git clone https://github.com/pritampanda15/PandaDock.git
cd PandaDock
pip install -e ".[gnn]"

This installs PandaDock with GNN support (PyTorch, PyTorch Geometric).

Basic Docking Example

The simplest way to perform molecular docking:

pandadock dock -r protein.pdb -l ligand.sdf \
               --center 10 20 30 --box 20 20 20 \
               -o results/

Where:

-r protein.pdb: Your receptor (protein) structure
-l ligand.sdf: Your ligand structure
--center X Y Z: Center coordinates of the binding site (Å)
--box X Y Z: Size of the docking grid box (Å)
-o results/: Output directory

Hybrid Docking with GNN (Recommended)

For best accuracy, use the hybrid workflow that combines traditional docking with GNN rescoring:

pandadock hybrid -r protein.pdb -l ligand.sdf \
                 --center 10 20 30 --box 20 20 20 \
                 --model models/best_model.pt \
                 -o hybrid_results/

This workflow:

Generates diverse poses using hierarchical docking
Rescores all poses using the SE(3)-equivariant GNN
Ranks poses by GNN-predicted binding affinity (pEC50)

GNN-Only Prediction

To predict binding affinity for an existing complex:

pandadock gnn predict -m models/best_model.pt \
                      -p protein.mol2 -l ligand.mol2

Training Your Own GNN Model

Train on the ULVSH dataset:

pandadock gnn train --dataset ULVSH/ --output models/ --epochs 100

Compare against baselines:

pandadock gnn compare --model models/best_model.pt \
                      --dataset ULVSH/ --output results/

Viewing Results

After docking completes, check the output directory:

complex1.pdb: Top-ranked protein-ligand complex
pose1.pdb: Top-ranked ligand pose
hybrid_results.csv: Complete results with GNN pEC50 values
interaction_analysis.json: Detailed interactions

Available Commands

Main docking commands:

pandadock dock: Traditional docking with Vina-style scoring
pandadock hybrid: Hybrid docking with GNN rescoring (recommended)
pandadock list-algorithms: Show available algorithms

GNN commands:

pandadock gnn download-model: Download pre-trained model
pandadock gnn train: Train GNN model
pandadock gnn predict: Predict binding affinity
pandadock gnn rescore: Rescore poses from any docking tool
pandadock gnn benchmark: Benchmark model performance
pandadock gnn compare: Compare against baselines

Other tools:

pandadock-prepare: Prepare ligands for docking
pandadock-gridbox: Generate grid box configurations
pandadock-flex: Flexible/induced-fit docking
pandadock-metal: Metal coordination docking
pandadock-tethered: Tethered docking for validation
pandadock-report: Generate publication figures

Next Steps

PandaDock-GNN Overview: Learn about the GNN architecture
../tutorials/gnn_training: Train your own model
../tutorials/hybrid_workflow: Master the hybrid workflow