alphafold-database

Retrieve AlphaFold protein structures and confidence metrics from UniProt accessions.

21|2|Updated Dec 8, 2025

Search Tags:#bioinformatics #pLDDT #drug discovery #PAE #alphafold #uniprot #protein structures

Ready to run

npx skills add https://github.com/silverstein/claude-scientific-skills-desktop --skill alphafold-database-silverstein

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: alphafold-database
Source: https://github.com/silverstein/claude-scientific-skills-desktop/tree/main/corpus/alphafold-database
Command: npx skills add https://github.com/silverstein/claude-scientific-skills-desktop --skill alphafold-database-silverstein

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps researchers retrieve AI-predicted protein structures and interpret their reliability without manually searching multiple sources or figuring out which files contain which confidence metrics.

Core Features & Use Cases

Predictive structure retrieval by identifier: Pull AlphaFold prediction records by UniProt accession and obtain the corresponding structure model files.
Confidence-aware structural analysis: Download and use pLDDT (per-residue confidence) and PAE (predicted aligned error) to judge which regions and relative domain arrangements are trustworthy.
Bulk dataset access for scaling up: Access proteome-scale datasets and metadata through Google Cloud and BigQuery for high-throughput workflows.
Practical structural biology and drug discovery workflows: Support downstream parsing, visualization, and comparison against experimental structures for protein engineering and structure-based studies.

Quick Start

Use the alphafold-database skill to retrieve AlphaFold predicted structure files and confidence metrics for UniProt accession P00520.

Frequently Asked Questions about alphafold-database

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I retrieve AlphaFold protein structures using a UniProt accession?▼

You retrieve AlphaFold protein structures by providing a UniProt accession identifier to fetch the corresponding predicted mmCIF or PDB model files for downstream analysis.

What is pLDDT and PAE in AlphaFold predicted structures?▼

pLDDT provides per-residue confidence scores to judge which protein regions are trustworthy, while the PAE matrix indicates predicted aligned error for evaluating relative domain arrangement confidence.

How do I access AlphaFold proteome-scale datasets in bulk?▼

You can access AlphaFold proteome-scale datasets in bulk through Google Cloud public datasets and BigQuery, utilizing cloud-based data sources with caching to enable efficient high-throughput scaling.

Does this AlphaFold retrieval method work for structure-based drug discovery?▼

Yes, retrieving AlphaFold predicted structures and confidence metrics directly supports practical structural biology and drug discovery workflows by enabling downstream parsing, visualization, and comparison against experimental structures.

What confidence metrics are included when downloading AlphaFold models?▼

Downloading AlphaFold models includes per-residue confidence files and PAE JSON files, allowing you to properly interpret which specific regions and relative domain arrangements are structurally reliable.

What's the best way to interpret AlphaFold confidence scores for protein engineering?▼

The best way to interpret AlphaFold confidence scores for protein engineering is to evaluate pLDDT for trustworthy residues and the PAE matrix for domain arrangement confidence during downstream structural comparison.