diffdock

Dock small-molecule ligands into protein pockets with DiffDock-L and rank poses by confidence.

288|34|Updated Jul 6, 2026

Search Tags:#molecular docking #ligand binding #protein structures #DiffDock-L #science simulation

Ready to run

npx skills add https://github.com/PKU-YuanGroup/OpenAI4S --skill diffdock-pku-yuangroup

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: diffdock
Source: https://github.com/PKU-YuanGroup/OpenAI4S/tree/main/skills/diffdock
Command: npx skills add https://github.com/PKU-YuanGroup/OpenAI4S --skill diffdock-pku-yuangroup

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill automates the process of small-molecule binding pose prediction with DiffDock-L, enabling scientists to rapidly assess potential ligand interactions.

Core Features & Use Cases

Ligand Docking: Predict the 3D binding poses of small molecules in protein pockets.
No Search Box: Perform blind docking without a predefined search box.
Confidence Ranking: Generate ranked poses using a learned confidence model.
Use Case: A scientist seeking to evaluate potential binders for a drug target can use this Skill to perform docking experiments efficiently.

Quick Start

Dock a molecule into a protein with DiffDock using its SKILL: dock small-molecule binder into the given PDB.

Frequently Asked Questions about diffdock

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I predict small-molecule ligand binding poses in protein structures?▼

To predict small-molecule ligand binding poses in protein structures, you can automate molecular docking using the DiffDock-L algorithm. This approach generates ranked 3D binding poses within protein pockets based on a learned confidence model to assess ligand interactions.

Can I perform blind molecular docking without defining a search box?▼

Yes, you can perform blind molecular docking without a predefined search box. This Skill utilizes the DiffDock-L algorithm to predict small-molecule binding poses directly within protein pockets without requiring manual search space specification.

Do I need GPU resources for ligand docking with DiffDock-L?▼

Yes, you need access to GPU resources for efficient ligand docking with DiffDock-L. Running the DiffDock-L algorithm to predict 3D binding poses and rank possible binders requires substantial computational power provided by a GPU.

How does confidence ranking work for molecular docking results?▼

Confidence ranking for molecular docking works by utilizing a learned confidence model to evaluate predicted 3D binding poses. The DiffDock-L algorithm ranks possible small-molecule binders based on these calculated confidence scores to identify optimal ligand interactions.

What is the best way to evaluate potential binders for a drug target?▼

The best way to evaluate potential binders for a drug target is to use automated molecular docking with DiffDock-L. It predicts small-molecule binding poses in protein pockets and ranks possible binders using a confidence model, enabling efficient docking experiments.

How do I dock a small-molecule binder into a given PDB protein structure?▼

To dock a small-molecule binder into a given PDB protein structure, apply the DiffDock-L algorithm to perform blind docking. The Skill processes the PDB file to predict and rank 3D ligand binding poses within the protein pocket using a confidence model.