Which Python libraries are required for this skill?

The skill primarily utilizes Biopython, Requests, NumPy, and Matplotlib. For bulk data, the google-cloud-bigquery and gsutil packages are also recommended.

How do I retrieve a protein structure using this skill?

You can retrieve structures by providing a UniProt ID. The skill uses the AlphaFold DB API or Biopython to fetch metadata and download the coordinate files in mmCIF or PDB format.

Can I download structures for an entire species at once?

Yes, the skill provides methods to access bulk proteome datasets via Google Cloud Storage using taxonomy IDs, such as 9606 for humans or 83333 for E. coli.

What are pLDDT and PAE metrics?

pLDDT measures per-residue confidence (0-100), where scores above 90 are high confidence. PAE (Predicted Aligned Error) indicates the confidence in the relative positions of different domains within the protein.

AlphaFold Database Tools

Name: AlphaFold Database Tools
Author: K-Dense-AI

byK-Dense-AI

•

3,719

•

数据科学与机器学习

Accesses and analyzes over 200 million AI-predicted protein structures for bioinformatics and structural biology research.

This skill provides a comprehensive interface for interacting with the AlphaFold Database, a repository of AI-predicted 3D protein structures maintained by DeepMind and EMBL-EBI. It allows users to retrieve structural data by UniProt ID, download coordinate files in mmCIF or PDB formats, and evaluate prediction reliability through confidence metrics like pLDDT and Predicted Aligned Error (PAE). Designed for researchers and bioinformaticians, it facilitates the integration of high-quality protein models into computational workflows, drug discovery pipelines, and large-scale proteomic studies using both direct API access and Google Cloud Platform integration.

主要功能

01Extract and interpret pLDDT and PAE confidence metrics for model validation

023,719 GitHub stars

03Retrieve 3D protein structure predictions using UniProt IDs or protein names

04Integrate with Biopython for advanced structural analysis and parsing

05Download atomic coordinates in standard PDB and mmCIF formats

06Access bulk proteome datasets via Google Cloud Platform and BigQuery

使用场景

01Structure-Based Drug Discovery: Identifying binding sites and performing docking on proteins lacking experimental structures

02Structural Proteomics: Analyzing large sets of protein models to identify high-confidence structural domains

03Protein Engineering: Designing mutations by evaluating residue stability and domain arrangements via confidence scores

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add k-dense-ai/claude-scientific-skills alphafold-database

For use in Claude.ai and ChatGPT

Download Skill