Skills clinvar-database
🧬

clinvar-database

Low Risk 🌐 Network access⚙️ External commands

Query ClinVar Genetic Variant Database

Also available from: davila7

ClinVar contains critical variant classifications for clinical genetics, but accessing and interpreting this data requires specialized knowledge. This skill provides comprehensive guidance for querying ClinVar via web interface, API, or FTP downloads, with detailed interpretation guidelines for clinical significance classifications.

Supports: Claude Codex Code(CC)
📊 69 Adequate
1

Download the skill ZIP

2

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

3

Toggle on and start using

Test it

Using "clinvar-database". Find pathogenic BRCA1 variants with expert review

Expected outcome:

  • Found 127 pathogenic BRCA1 variants with three or four star review status
  • Top variants by clinical importance:
  • - NM_007294.3:c.68_69delAG (185delAG) - Founder mutation
  • - NM_007294.3:c.5266dupC (5382insC) - Common pathogenic variant
  • - NM_007294.3:c.181T>G (C61G) - High-penetrance mutation
  • All variants confirmed by ClinGen expert panel or practice guidelines

Security Audit

Low Risk
v4 • 1/17/2026

This is a documentation-only skill containing markdown files with examples for accessing NCBI's ClinVar database. All 241 static findings are FALSE POSITIVES triggered by code examples in documentation. The skill contains no executable code - only markdown documentation showing legitimate examples for querying a public genomic database. No actual security risks present.

5
Files scanned
1,418
Lines analyzed
2
findings
4
Total audits

Risk Factors

🌐 Network access (3)
SKILL.md:33 SKILL.md:54 references/api_reference.md:15
⚙️ External commands (3)
SKILL.md:52-55 SKILL.md:120-125 references/api_reference.md:37-39
Audited by: claude View Audit History →

Quality Score

45
Architecture
100
Maintainability
83
Content
30
Community
90
Security
83
Spec Compliance

What You Can Build

Identify High-Confidence Pathogenic Variants

Search for pathogenic variants in specific genes with expert panel review status for clinical reporting.

Annotate VCF with Clinical Significance

Download ClinVar VCF and annotate patient variant calls with clinical significance classifications.

Analyze Disease-Associated Variants

Study all variants associated with a specific condition and analyze their distribution across genes.

Try These Prompts

Basic Gene Search 
Search ClinVar for all variants in the BRCA1 gene and show me the pathogenic ones with high review status.
Clinical Interpretation 
What does Uncertain Significance mean in ClinVar, and how should I interpret conflicting classifications?
Bulk Data Download 
Show me how to download the latest ClinVar VCF file for GRCh38 and annotate my variants with it.
Advanced Query 
Find all likely pathogenic variants in CFTR that have been reviewed by expert panels, excluding conflicting interpretations.

Best Practices

  • Always check review status (star rating) - prefer three or four star classifications
  • Document ClinVar version and access date for reproducibility
  • Test queries on web interface before automating with API

Avoid

  • Using VUS variants for clinical decisions without expert review
  • Ignoring conflicting interpretations without manual evaluation
  • Assuming classifications are permanent - they update with new evidence

Frequently Asked Questions

What is the difference between Pathogenic and Likely Pathogenic?
Pathogenic means approximately 99 percent probability of causing disease, while Likely Pathogenic is approximately 90 percent. Both are clinically actionable but with different confidence levels.
How often is ClinVar updated?
The website updates weekly on Mondays, with monthly full releases on the first Thursday. Use monthly releases for reproducible research.
Can I use ClinVar for direct patient diagnosis?
No, ClinVar is a research database. All clinical decisions require interpretation by a qualified genetics professional.
What is the best format for bulk downloads?
XML files contain the most comprehensive data including all submissions and evidence. VCF files are best for genomic pipelines.
How do I handle conflicting interpretations?
Check review status (prefer three or four star), examine evidence from each submitter, consider submission dates, and consult expert panels when available.
Do I need an API key?
Not required but recommended. Without a key: 3 requests per second. With NCBI API key: 10 requests per second. Register at ncbi.nlm.nih.gov account.

Developer Details

Author

K-Dense-AI

License

MIT

Repository

https://github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/clinvar-database

Ref

main

File structure

📁 references/

📄 api_reference.md

📄 clinical_significance.md

📄 data_formats.md

📄 SKILL.md