SkillsAdd

mims-harvard/tooluniverse

tooluniverse-structural-proteomics

Structural biology plus proteomics integration for drug target validation.

First seen Mar 25, 2026

Installation

$ npx skills add mims-harvard/tooluniverse --skill tooluniverse-structural-proteomics
View on skills.sh ↗ GitHub repository ↗

Summary

  • Structural biology plus proteomics integration for drug target validation.
  • Combines PDB experimental structures, AlphaFold predictions, GPCRdb, SAbDab antibody structures, ProteinsPlus binding-site prediction, and BindingDB ligand-affinity data.
  • Use for druggability assessment, binding-site characterization, ligand-pocket analysis, structural-confidence scoring (resolution, pLDDT), and antibody-target interface analysis.

Also in this package

Other skills from mims-harvard/tooluniverse · top by installs.

npx skills add mims-harvard/tooluniverse
tooluniverse-sequence-retrieval 1.6K tooluniverse-literature-deep-research 643 tooluniverse-image-analysis 637 tooluniverse-clinical-guidelines 466 tooluniverse 458 tooluniverse-drug-research 452 tooluniverse-clinical-trial-design 440 tooluniverse-drug-drug-interaction 432 setup-tooluniverse 427 tooluniverse-disease-research 419

Browse all from mims-harvard/tooluniverse

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 1.7K
License LICENSE
Default branch main
Open issues 9
Status Active

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 6,521 B
  • docs SUMMARY.md 463 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 218 installs

SKILL.md

Structural Proteomics for Drug Target Validation

Comprehensive structural data integration using ToolUniverse tools across PDB, AlphaFold, GPCRdb, SAbDab, and proteomics databases for drug target validation.

LOOK UP DON'T GUESS

  • PDB structures/resolutions: PDBeSIFTSgetbeststructures and RCSBGraphQLgetstructuresummary
  • AlphaFold confidence: alphafoldgetsummary
  • Ligands/affinities: PDBegetstructureligands and BindingDBgetligandsby_uniprot
  • Druggability: ProteinsPluspredictbinding_sites

COMPUTE, DON'T DESCRIBE

When analysis requires computation (statistics, data processing, scoring, enrichment), write and run Python code via Bash. Don't describe what you would do — execute it and report actual results. Use ToolUniverse tools to retrieve data, then Python (pandas, scipy, statsmodels, matplotlib) to analyze it.

Domain Reasoning

Resolution determines valid conclusions: <2A = atom positions visible; 2-3A = side chains reliable, drug design supported; >3A = backbone only, binding site unreliable. Do not over-interpret low-resolution structures.

Tool Inventory

PDB (RCSB)

RCSBAdvSearchsearchstructures (querytype, queryvalue, rows), RCSBDatagetentry (entryid), RCSBGraphQLgetstructuresummary (pdbid), RCSBGraphQLgetligandinfo (pdbid), RCSBgetchemicalcomponent (comp_id)

PDB (PDBe)

pdbegetentrysummary (pdbid), PDBegetstructureligands (pdbid), PDBegetboundmolecules (pdbid), PDBeSearchsearchstructures (query, rows), PDBeSIFTSgetbeststructures (uniprotid), PDBeSIFTSgetallstructures (uniprotid), PDBeKBgetligandsites (pdbid), PDBeKBgetinterfaceresidues (pdbid), PDBeValidationgetqualityscores (pdbid)

PDBe PISA

PDBePISAgetinterfaces (pdbid), PDBePISAgetassemblies (pdbid)

AlphaFold

alphafoldgetprediction (qualifier=UniProt), alphafoldgetsummary (qualifier), alphafoldgetannotations (qualifier)

Binding Sites

ProteinsPluspredictbindingsites (pdbid, chain), BindingDBgetligandsbyuniprot (uniprotid), BindingDBgetligandsbypdb (pdbid), BindingDBgettargetsbycompound (smiles)

Foldseek

Foldseeksearchstructure (sequence, mode="tmalign"), Foldseekgetresult (ticket)

GPCRdb

GPCRdbgetprotein (protein), GPCRdbgetstructures (protein), GPCRdbgetligands (protein), GPCRdbgetmutations (protein). Accepts entry names, gene symbols (auto-converted to {symbol.lower()}_human), or UniProt accessions.

SAbDab

SAbDabsearchstructures (query/antigen), SAbDabgetstructure (pdbid), TheraSAbDabsearchtherapeutics (query), TheraSAbDabsearchbytarget (target)

Domains

InterProgetproteindomains (uniprotid), Pfamgetproteinannotations (uniprotid), UniProtgetentrybyaccession (accession)

Proteomics

ProteomeXchangesearchdatasets (query), ProteomeXchangegetdataset (dataset_id)

Workflow 1: Find All Structures for a Drug Target

Phase 0: Resolve protein → UniProt ID, gene symbol, organism
Phase 1: PDBeSIFTS_get_best_structures → RCSBGraphQL_get_structure_summary → PDBeValidation
Phase 2: alphafold_get_prediction/summary → compare pLDDT with experimental coverage
Phase 3: IF GPCR → GPCRdb; IF antibody target → SAbDab/TheraSAbDab
Phase 4: InterPro/Pfam domain mapping → identify unresolved regions
Phase 5: Summary table (PDB ID, method, resolution, ligands, coverage, quality)

Decisions: Resolution <2.5A for drug design. X-ray > Cryo-EM > NMR > AlphaFold for binding sites. Holo > apo structures.

Workflow 2: Identify Binding Pocket Ligands

Phase 1: PDBe_get_structure_ligands + RCSBGraphQL_get_ligand_info + PDBe_KB_get_ligand_sites
Phase 2: ProteinsPlus_predict_binding_sites → druggability score, pocket residues
Phase 3: BindingDB_get_ligands_by_pdb/uniprot → Ki, Kd, IC50
Phase 4: RCSB_get_chemical_component for key ligands

Filter artifacts: GOL, EDO, SO4, PEG, ACT, CL, NA. Keep cofactors (ATP, NAD, HEM) and catalytic metals (ZN, MG) if relevant.

Workflow 3: Cross-Validate Drug Binding

Phase 1: Find co-crystal structures → filter for drug/analogs
Phase 2: BindingDB affinity data (Ki, Kd, IC50)
Phase 3: ProteinsPlus + PDBe-KB binding site characterization
Phase 4: PDBeValidation quality → binding site well-resolved?
Phase 5: AlphaFold + Foldseek structural comparison
Phase 6: GPCR-specific (if applicable) → active/inactive states, pharmacology, resistance mutations
Phase 7: Antibody-specific (if applicable) → epitope mapping
Phase 8: Evidence integration

Tool Parameter Gotchas

Tool Mistake Correct
alphafoldgetprediction/summary uniprot_id qualifier
GPCRdbgetprotein gene_name protein
PDBeSIFTSgetbest_structures gene symbol uniprot_id (e.g., "P04637")
Foldseeksearchstructure mode="3diaa" mode="tmalign"
SAbDabsearchstructures name query or antigen
RCSBgetchemical_component ligand_id comp_id

Evidence Grading

Tier Confidence
T1 Co-crystal (<2.5A) + binding affinity data
T2 Experimental structure + computational prediction
T3 AlphaFold + pocket analysis + known ligand analogs
T4 Homology model or low-resolution only

Interpretation

Metric High Acceptable Caution
Resolution <2.0A (X-ray) / <3.0A (cryo-EM) 2.0-2.5A / 3.0-4.0A >3.0A / >4.5A
R-free <0.25 0.25-0.30 >0.30
AlphaFold pLDDT >90 70-90 <70 (disordered)

DoGSiteScorer >0.6 = druggable; <0.4 = unlikely druggable. PISA assemblies should be cross-validated with SEC-MALS/native MS.

Limitations

  • BindingDB: 60s+ for popular targets
  • AlphaFold: lacks ligand context
  • GPCRdb: Class A-F GPCRs only
  • PDBePISA: operation is internal, not a public parameter