Is a GPU required to run this skill effectively?

While it can run on a CPU, a GPU is strongly recommended as it typically offers a 10-100x speedup for computational docking tasks.

Does this skill support batch virtual screening?

Yes, it includes specific workflows for batch processing multiple complexes using CSV input files and pre-computed protein embeddings for efficiency.

Can DiffDock predict binding affinity?

No, DiffDock focuses on predicting 3D binding poses and confidence scores. It should be combined with specialized scoring functions like GNINA or MM/GBSA for affinity calculations.

What inputs are required for the DiffDock skill?

You need a protein target (either a PDB file or amino acid sequence) and a ligand description (SMILES string, SDF, or MOL2 file).

How should I interpret the generated confidence scores?

Scores above 0 indicate high confidence, -1.5 to 0 are moderate, and scores below -1.5 suggest low prediction certainty that require careful validation.

DiffDock Molecular Docking

Name: DiffDock Molecular Docking
Author: Microck

byMicrock

•

데이터 과학 및 ML

Predicts 3D protein-ligand binding poses using state-of-the-art diffusion models for structure-based drug discovery.

소개

DiffDock is a specialized Claude Code Skill that enables researchers and developers to perform diffusion-based molecular docking directly within their workflow. By leveraging deep learning, it predicts highly accurate 3D binding poses of small molecules to protein targets, supporting both structure-based inputs (PDB) and sequence-based inputs (ESMFold). This skill is essential for virtual screening, lead optimization, and structure-based drug design, providing confidence scores to validate prediction reliability while integrating seamlessly with downstream scoring functions for affinity assessment.

주요 기능

Handles diverse ligand inputs including SMILES, SDF, and MOL2 formats
Generates confidence scores to assess 3D pose prediction reliability
Predicts ligand binding poses with high-accuracy diffusion models
Batch processing capabilities for large-scale virtual screening campaigns
81 GitHub stars
Supports protein structures (PDB) and amino acid sequences via ESMFold

사용 사례

Predicting binding sites for novel small molecules during lead optimization
Virtual screening of compound libraries against specific protein targets
Automating molecular docking workflows for structure-based drug design

소개

주요 기능

Handles diverse ligand inputs including SMILES, SDF, and MOL2 formats
Generates confidence scores to assess 3D pose prediction reliability
Predicts ligand binding poses with high-accuracy diffusion models
Batch processing capabilities for large-scale virtual screening campaigns
81 GitHub stars
Supports protein structures (PDB) and amino acid sequences via ESMFold

사용 사례

Predicting binding sites for novel small molecules during lead optimization
Virtual screening of compound libraries against specific protein targets
Automating molecular docking workflows for structure-based drug design