What file formats are supported for downloads?

The skill supports downloading model coordinates in mmCIF/PDBx and PDB formats, as well as JSON files for confidence scores and Predicted Aligned Error (PAE) matrices.

Can I process multiple proteins at once?

Yes, the skill provides patterns for batch processing UniProt lists and downloading entire proteomes via Google Cloud Storage and BigQuery metadata.

What are pLDDT scores and why do they matter?

pLDDT measures AlphaFold's per-residue confidence. Scores above 90 indicate high accuracy suitable for detailed structural analysis, while scores below 50 suggest disordered regions.

Does this skill require a local installation of AlphaFold?

No, it interacts with the AlphaFold DB API and public datasets, allowing you to access pre-calculated predictions without running the full AlphaFold model locally.

How do I find a protein structure using this skill?

You can search by UniProt accession ID or protein name to retrieve metadata and direct download links for the corresponding AlphaFold prediction.

AlphaFold Protein Structure Database

Name: AlphaFold Protein Structure Database
Author: jimmc414

byjimmc414

•

324

データサイエンスとML

Access and analyze over 200 million AI-predicted protein structures for structural biology and computational drug discovery.

概要

The AlphaFold Database skill empowers Claude to interact with the world's most comprehensive repository of predicted protein structures. It enables researchers and developers to retrieve high-accuracy 3D models using UniProt IDs, download coordinate files in mmCIF or PDB formats, and interpret critical confidence metrics like pLDDT and Predicted Aligned Error (PAE). Designed for structural proteomics and bio-computation, this skill streamlines the integration of AlphaFold's massive dataset into automated workflows, facilitating everything from protein engineering to bulk proteome analysis via the EBI API and Google Cloud Platform.

主な機能

Integration with Biopython for programmatic structure parsing and analysis
Retrieve 3D protein structures using UniProt accession IDs or protein names
Bulk data access capabilities via Google Cloud Platform and BigQuery
324 GitHub stars
Download atomic coordinates in standard PDB and mmCIF file formats
Access and interpret pLDDT and PAE confidence metrics for reliability assessment

ユースケース

Identifying potential drug binding sites and performing structure-based docking studies
Analyzing domain architectures and residue-level stability for protein engineering
Building structural models for proteins lacking experimental data from the PDB

概要

主な機能

Integration with Biopython for programmatic structure parsing and analysis
Retrieve 3D protein structures using UniProt accession IDs or protein names
Bulk data access capabilities via Google Cloud Platform and BigQuery
324 GitHub stars
Download atomic coordinates in standard PDB and mmCIF file formats
Access and interpret pLDDT and PAE confidence metrics for reliability assessment

ユースケース

Identifying potential drug binding sites and performing structure-based docking studies
Analyzing domain architectures and residue-level stability for protein engineering
Building structural models for proteins lacking experimental data from the PDB