Claude Skill MartBrowse skillsWhat is a Skill?
 
← Back to list

DiffDock Molecular Docking

Runs DiffDock/DiffDock-L to predict protein–small-molecule binding poses and helps you interpret the confidence scores.

Data & AnalyticsAdvanced★ 33,030⑂ 3,248AI score 8/10Last updated: Aug 9, 2026

What it does

  • Builds correct DiffDock inference commands from a PDB file (or amino-acid sequence) plus a SMILES string, SDF, or MOL2 ligand.
  • Covers single-complex docking, CSV-driven batch docking, and large virtual screening with pre-computed ESM embeddings.
  • Parses rank*_confidence*.sdf outputs, classifies confidence as High/Moderate/Low, ranks poses, and exports CSV summaries.
  • Tunes key parameters (samples_per_complex, inference_steps, torsion temperature) using presets for high accuracy, fast screening, flexible or rigid ligands.
  • Explains downstream rescoring with GNINA, MM/GBSA, or FEP, and is explicit about scope limits (no affinity prediction, no protein–protein docking).

Who it's for

  • Computational chemists and chemical biologists doing structure-based drug design and lead optimization.
  • Biotech/pharma teams screening compound libraries against a target.
  • Intermediate-to-advanced users comfortable with Conda, PyTorch/PyG, and a CUDA GPU.

Examples

  1. "Dock this aspirin SMILES into protein.pdb" → generates the python -m inference --protein_path ... --ligand_description ... call and interprets ranked poses.
  2. "Screen 200 compounds against one target" → creates and validates the batch CSV, pre-computes ESM embeddings, runs batch inference, exports the top 20 poses.
  3. "All my poses score around -2 confidence, why?" → diagnoses likely causes (bulky ligand, flexible protein) and suggests more samples or ensemble docking.

· · · Install guide · · ·

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/diffdock folder from the GitHub repo K-Dense-AI/scientific-agent-skills into my ~/.claude/skills/diffdock/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/K-Dense-AI/scientific-agent-skills.git && mkdir -p ~/.claude/skills && cp -r scientific-agent-skills/skills/diffdock ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Install DiffDock itself first: git clone https://github.com/gcorso/DiffDock.git, then cd DiffDock && conda env create --file environment.yml && conda activate diffdock (or docker pull rbgcsail/diffdock).
  2. Verify GPU access: python -c "import torch; print(torch.cuda.is_available())" should print True for practical runtimes.
  3. Clone the skill repo: git clone https://github.com/K-Dense-AI/scientific-agent-skills.git
  4. Copy the skill into place: mkdir -p ~/.claude/skills && cp -r scientific-agent-skills/skills/diffdock ~/.claude/skills/
  5. Restart Claude Code and start it from inside your DiffDock checkout.
  6. Ask "check my DiffDock environment" so the skill runs scripts/setup_check.py and reports missing dependencies.
  7. Then just say things like "dock this ligand to this protein" and the skill triggers automatically.
View source on GitHub ↗License: MIT