Does DiffDock predict binding affinity?

No, DiffDock is specifically designed to predict 3D binding poses and provide confidence scores for those poses. It should be combined with scoring functions like GNINA or MM/GBSA for affinity assessment.

Can I use amino acid sequences instead of PDB files?

Yes, DiffDock supports protein sequences by utilizing ESMFold to generate structural predictions, which it then uses for the docking process.

Is a GPU required to run DiffDock?

A GPU is strongly recommended for practical use, as it offers a 10-100x speedup compared to CPU processing, making it feasible for batch screening tasks.

What is the significance of the confidence score?

Confidence scores indicate the model's certainty about the structural pose. Scores above 0 are high confidence, while scores below -1.5 suggest the prediction is uncertain and requires careful validation.

DiffDock: Molecular Docking & Drug Design

Name: DiffDock: Molecular Docking & Drug Design
Author: henriquescastilho

byhenriquescastilho

•

데이터 과학 및 ML

Predicts accurate protein-ligand binding poses using diffusion-based deep learning models for structure-based drug discovery.

DiffDock is a specialized skill for Claude Code that enables high-accuracy molecular docking through state-of-the-art diffusion models. It allows researchers and computational biologists to predict the 3D orientation of small molecule ligands within protein binding sites, generating ranked poses with associated confidence scores. Whether processing single complexes or running large-scale virtual screening campaigns, DiffDock streamlines the transition from chemical strings and PDB files to actionable structural insights, serving as a critical component in the lead optimization and drug discovery pipeline.

주요 기능

01Generates confidence scores to assess prediction reliability

021 GitHub stars

03Predicts high-accuracy 3D ligand binding poses using diffusion models

04Supports both protein structures (PDB) and sequences (via ESMFold)

05Handles batch processing for large virtual screening campaigns

06Integrates with diverse chemical formats including SMILES, SDF, and MOL2

사용 사례

01Modeling binding poses for proteins where only the amino acid sequence is known

02Screening a library of compounds against a specific protein target

03Predicting the binding site and orientation of a novel drug candidate

주요 기능

01Generates confidence scores to assess prediction reliability

021 GitHub stars

03Predicts high-accuracy 3D ligand binding poses using diffusion models

04Supports both protein structures (PDB) and sequences (via ESMFold)

05Handles batch processing for large virtual screening campaigns

06Integrates with diverse chemical formats including SMILES, SDF, and MOL2

사용 사례

01Modeling binding poses for proteins where only the amino acid sequence is known

02Screening a library of compounds against a specific protein target

03Predicting the binding site and orientation of a novel drug candidate