Diffdock

This is a diffusion-based deep learning model for predicting how small molecules bind to proteins in 3D, which matters if you're doing drug discovery or chemical biology work. You give it a protein structure (PDB or sequence via ESMFold) and a ligand (SMILES or structure file), and it generates ranked binding poses with confidence scores. The key thing to understand is it predicts pose geometry and model certainty, not binding affinity, so you'll need to pair it with scoring functions like GNINA for energy estimates. It handles batch virtual screening well, especially if you pre-compute protein embeddings for large libraries. GPU recommended since it's 10-100x faster than CPU, and the first run takes a few minutes to build lookup tables.

skills.sh installs: 2K
GitHub stars: 56.6K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add k-dense-ai/scientific-agent-skills --skill diffdock --agent claude-code

Installs into .claude/skills of the current project.

Files

DiffDock: protein-small-molecule docking

DiffDock-L generates candidate ligand poses and ranks them by model confidence. Confidence is neither a measured probability of correctness nor binding affinity. Use this skill for one pair, batches, or separate receptor conformations; treat library-wide confidence sorting as pose triage, not hit identification.

Verified scope

Targets the current v1.1.3 release. Released source and current main's identical inference.py were checked on 2026-09-30. Bundled helpers were tested on tiny synthetic inputs; pretrained docking, ESMFold, CUDA, Docker, GNINA, and hosted-demo execution were not run in this review. Commands requiring those components are source-verified recipes, not successful end-to-end demonstrations.

Set up the upstream environment

git clone --branch v1.1.3 --depth 1 https://github.com/gcorso/DiffDock.git
cd DiffDock
conda env create --file environment.yml
conda activate diffdock

SKILL.md on GitHub