skills.rest
mims-harvard
mims-harvardOFFICIAL
Explore all skills by this author →
Website𝕏@marinkazitnikGitHub

tooluniverse-structural-proteomics

Integrates PDB, AlphaFold, GPCRdb, and BindingDB data for drug target structural validation.

1.7k|254|Updated Mar 3, 2025
Search Tags:#pdb#proteomics#alphafold#druggability#structural biology#binding site#drug target validation
One-click install
npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-structural-proteomics
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tooluniverse-structural-proteomics
Source: https://github.com/mims-harvard/ToolUniverse/tree/main/plugins/tooluniverse/skills/tooluniverse-structural-proteomics
Command: npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-structural-proteomics

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Validating a drug target requires pulling structural evidence from many disconnected databases (PDB, AlphaFold, GPCRdb, SAbDab, BindingDB) and judging whether the data is trustworthy enough to support drug design decisions. This Skill orchestrates those lookups and applies resolution, pLDDT, and druggability criteria so conclusions are grounded in actual structural evidence.

Core Features & Use Cases

  • Structure Discovery: Find all experimental and predicted structures for a target via PDBeSIFTS, RCSB, and AlphaFold, with quality scoring from PDBeValidation.
  • Binding-Site & Ligand Analysis: Characterize pockets with ProteinsPlus, retrieve ligands and affinities (Ki, Kd, IC50) from BindingDB, and filter crystallization artifacts.
  • Specialized Targets: Query GPCRdb for GPCR states and mutations, and SAbDab/TheraSAbDab for antibody-target interfaces and epitope mapping.
  • Use Case: Given a UniProt ID for a kinase target, compile a summary table of all PDB structures with resolution, ligands, and coverage, then grade the druggability evidence from T1 (co-crystal plus affinity) to T4 (homology model only).

Quick Start

Find all available structures for UniProt target P04637, assess their resolution and quality, and evaluate the druggability of its binding pockets.

Frequently Asked Questions about tooluniverse-structural-proteomics

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

FAQPage Schema
How do I find all PDB structures for a drug target?▼

Start with PDBeSIFTS_get_best_structures using the UniProt ID, then enrich each hit with RCSBGraphQL_get_structure_summary and PDBeValidation_get_quality_scores. Complement with alphafold_get_prediction to compare pLDDT confidence against experimental coverage.

How to assess binding pocket druggability from a PDB structure?▼

Run ProteinsPlus_predict_binding_sites on the PDB ID to get pocket residues and a DoGSiteScorer druggability score. Scores above 0.6 indicate druggable pockets, while scores below 0.4 suggest the site is unlikely to be druggable.

What resolution is needed for structure-based drug design?▼

Structures below 2.5A resolution are recommended for drug design, with under 2.0A being high confidence for X-ray data. Above 3.0A only the backbone is reliable, so binding-site conclusions from such structures should be treated with caution.

Does AlphaFold confidence replace experimental structures?▼

No. AlphaFold predictions lack ligand context and should be cross-validated with experimental data. pLDDT above 90 is high confidence, 70-90 is acceptable, and below 70 typically indicates disordered regions unsuitable for binding-site analysis.

What are the limitations of GPCRdb and BindingDB queries?▼

GPCRdb only covers Class A-F GPCRs and accepts entry names, gene symbols, or UniProt accessions via the protein parameter. BindingDB queries for popular targets can take over 60 seconds to return ligand affinity data.

Related Skills

Explore More Education & Research Skills→
Runchuan-BURunchuan-BU
401

query-uniprot

Retrieve protein sequences and functional annotations from the UniProt REST API.

Community
Intermediate
jaechang-hitsjaechang-hits
298

autodock-vina-docking

Perform molecular docking with AutoDock Vina to predict protein-ligand binding poses and affinities.

Community
Advanced
monarch-initiativemonarch-initiative
18

gene-set-enrichment

Identify enriched pathways, GO terms, and disease associations from gene lists.

Official
Intermediate
pradyumnasagarpradyumnasagar
6

docking-rescoring

Predict and score protein-ligand binding affinities with machine learning and docking.

Community
Advanced