How can I tell if an AlphaFold predicted structure is reliable?

The skill provides patterns to analyze pLDDT scores (per-residue confidence) and PAE (Predicted Aligned Error) matrices to verify structure reliability.

What file formats are available for protein models?

The skill supports downloading atomic coordinates in both mmCIF (PDBx) and PDB formats, along with JSON-formatted confidence metrics.

Does this skill support bulk data downloads for entire species?

Yes, it includes specialized commands for accessing complete proteomes and metadata via Google Cloud Storage and BigQuery.

How do I fetch a protein structure using this skill?

You can retrieve structures by UniProt ID using Biopython or direct REST API calls as demonstrated in the skill's core capabilities section.

AlphaFold Database Access

Name: AlphaFold Database Access
Author: lifangda

bylifangda

•

データサイエンスとML

Facilitates the retrieval and analysis of over 200 million AI-predicted protein structures from the AlphaFold DB for biological research and drug discovery.

This skill empowers Claude to interact with the AlphaFold Database, a massive repository of high-accuracy protein structure predictions maintained by DeepMind and EMBL-EBI. It provides specialized workflows for fetching coordinate files (PDB/mmCIF), evaluating prediction reliability through pLDDT and PAE metrics, and performing bulk downloads via Google Cloud Platform. By integrating biological data science libraries like Biopython and structural analysis tools, it enables researchers and developers to automate the process of building structural models, identifying druggable pockets, and conducting large-scale proteomic studies directly within their terminal environment.

主な機能

01Access large-scale proteome datasets via Google Cloud BigQuery and GCS integration

02Automate structure parsing and contact mapping using Biopython

03Retrieve 3D protein structures by UniProt ID or protein name using REST APIs

04Download high-quality PDB and mmCIF coordinate files for structural modeling

05Analyze pLDDT and PAE metrics to assess prediction confidence and domain stability

0616 GitHub stars

ユースケース

01Automating structural proteomics workflows for proteins lacking experimental data

02Protein engineering and mutation analysis using predicted stability and confidence scores

03Structure-based drug discovery and binding site identification for novel targets

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lifangda/claude-plugins alphafold-database

For use in Claude.ai and ChatGPT

Download Skill