| name | protein-structure-analysis |
| description | Access AlphaFold Protein Structure Database and Protein Data Bank (PDB) for structural biology research. Analyze protein structures and molecular interactions.
|
Protein Structure Database (AlphaFold & PDB)
Access the AlphaFold Protein Structure Database and Protein Data Bank (PDB) for structural biology research, drug target analysis, and molecular interaction studies. This skill enables querying protein structures by gene name, UniProt accession, or PDB ID, retrieving 3D structural data, analyzing binding sites and protein-ligand interactions, and comparing predicted (AlphaFold) and experimentally determined (PDB) structures.
Quick Install
npx skills add Open-Medica/open-medical-skills --skill protein-structure-analysis
What It Does
- Structure retrieval: Queries AlphaFold DB and PDB by UniProt accession number, gene symbol, protein name, or PDB ID to retrieve atomic coordinates, confidence scores (pLDDT for AlphaFold), resolution data (for experimental structures), and associated metadata
- Binding site analysis: Identifies known ligand binding sites, catalytic residues, and allosteric pockets from PDB co-crystal structures, mapping residue-level interactions (hydrogen bonds, hydrophobic contacts, pi-stacking, salt bridges) between proteins and bound molecules
- Structure comparison: Aligns predicted AlphaFold models against experimentally determined PDB structures to assess prediction accuracy, identify regions of high uncertainty (low pLDDT / high PAE), and evaluate whether the predicted structure is suitable for downstream computational studies
- Variant impact mapping: Maps clinically relevant amino acid variants onto 3D structures to assess whether mutations occur in functionally important regions (active sites, protein-protein interfaces, transmembrane domains, signal peptides) that could explain pathogenicity
- Drug target assessment: Evaluates protein structures for druggability by analyzing surface pocket geometry, hydrophobicity, and comparison with known drug target families, supporting early-stage drug discovery and repurposing efforts
Clinical Use Cases
- Pharmacology target validation: A research team investigating a novel kinase inhibitor retrieves the target kinase structure from PDB, analyzes the ATP-binding pocket geometry, and compares it against the binding modes of existing approved inhibitors to guide lead optimization
- Variant of uncertain significance interpretation: A clinical geneticist evaluating a missense VUS in a cardiac ion channel protein uses the skill to map the affected residue onto the AlphaFold-predicted structure, determining that it falls within the selectivity filter region — a critical functional domain — supporting reclassification toward likely pathogenic
- Antibiotic resistance mechanism analysis: A microbiologist investigates a resistance-conferring mutation in a bacterial ribosomal protein by retrieving the ribosome structure from PDB and analyzing how the mutation disrupts the antibiotic binding site
- Protein engineering: A biotechnology team designing a modified therapeutic antibody uses the skill to analyze the antigen-binding interface, identify critical contact residues, and model how proposed mutations might affect binding affinity
Safety & Evidence
- Safety Classification: Safe — The skill retrieves and analyzes publicly available structural biology data from established scientific databases. It does not generate clinical recommendations, modify patient data, or influence treatment decisions directly. Structural interpretations for clinical variant assessment must be validated by molecular pathologists or structural biologists.
- Evidence Level: High — AlphaFold predictions are generated by DeepMind's AlphaFold2/AlphaFold3 system, which achieved near-experimental accuracy in CASP14/CASP15 competitions. PDB structures are experimentally determined by X-ray crystallography, cryo-EM, or NMR spectroscopy and are subject to peer review. Both databases are maintained by leading international scientific organizations (EMBL-EBI, wwPDB).
Example Usage
Protein structure query and analysis:
Retrieve and analyze the structure of human ACE2 receptor:
- UniProt: Q9BYF1
- Show binding interface with SARS-CoV-2 spike protein RBD
- Identify key contact residues
ACE2 Receptor Structure Analysis:
AlphaFold predicted structure:
- UniProt: Q9BYF1 | Gene: ACE2 | Length: 805 residues
- Overall pLDDT: 91.3 (Very High Confidence)
- Low-confidence regions: Residues 697-716 (transmembrane domain, pLDDT 52-68)
PDB experimental structures available:
- 6M0J: ACE2 bound to SARS-CoV-2 RBD (2.45 A resolution, X-ray)
- 6VW1: ACE2 bound to SARS-CoV-2 RBD chimera (2.68 A resolution)
- 7A94: Full-length ACE2 with B0AT1 (2.90 A, cryo-EM)
Key ACE2-RBD interface contacts (from PDB 6M0J):
| ACE2 Residue | RBD Residue | Interaction Type |
|---|
| K31 | Q493 | Hydrogen bond |
| E35 | Q493 | Hydrogen bond |
| D38 | Y449 | Hydrogen bond |
| Y41 | T500 | Hydrogen bond |
| Q42 | G446, Y449 | Hydrogen bonds |
| K353 | G502, G496 | Hydrogen bond network |
| D30 | K417 | Salt bridge |
Druggable pockets identified: Catalytic zinc-binding site (HEXXH motif at residues 374-378), substrate binding cleft — potential target for small molecule modulators of ACE2-RBD interaction.
Technical Details
- Category: Clinical Research Summarizing
- Author: OMS Contributors
- License: MIT
- Version: 1.0.0
- Specialty: Biochemistry, Molecular Biology
References
- Jumper J, et al. "Highly accurate protein structure prediction with AlphaFold." Nature. 2021;596:583-589.
- Berman HM, et al. "The Protein Data Bank." Nucleic Acids Res. 2000;28(1):235-242.
- Varadi M, et al. "AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models." Nucleic Acids Res. 2022;50(D1):D431-D437.
- Lan J, et al. "Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor." Nature. 2020;581:215-220.
- wwPDB consortium. "Protein Data Bank: the single global archive for 3D macromolecular structure data." Nucleic Acids Res. 2019;47(D1):D520-D528.
This skill is part of Open Medical Skills, a curated marketplace of medical AI skills maintained by physicians for physicians and the healthcare industry.
