zinc-database
Search ZINC database for compounds
Also available from: davila7
The ZINC database contains 230M+ purchasable chemical compounds for drug discovery research. This skill provides programmatic access to search by ZINC ID, SMILES, or supplier code, retrieve 3D structures, and perform similarity searches for virtual screening and molecular docking studies.
Download the skill ZIP
Upload in Claude
Go to Settings → Capabilities → Skills → Upload skill
Toggle on and start using
Test it
Using "zinc-database". Search for ZINC000000000001 and find similar compounds with distance 3
Expected outcome:
- ZINC ID: ZINC000000000001
- SMILES: [displayed structure]
- Available from: [catalogs]
- Similar compounds found: 142
- Top matches: ZINC000000000002, ZINC000000000003, ...
Using "zinc-database". Get 100 random lead-like compounds with MW 250-350
Expected outcome:
- Retrieved 100 compounds from ZINC22
- Property range: MW 250-350, lead-like subset
- Tranche codes: H##P###M### format
- Each entry includes ZINC ID, SMILES, and supplier information
Security Audit
SafeAll 389 static findings are false positives. The skill is legitimate scientific documentation for accessing the public ZINC database maintained by UCSF. subprocess.run calls use array syntax with hardcoded ZINC API endpoints. C2 keyword flags are common programming terms (command, control, execute) used in legitimate documentation context. The scanner misinterpreted cheminformatics terminology (MD5/SHA checksums) and DOCK database format files (.db2) as security concerns. No malicious behavior patterns detected.
Risk Factors
🌐 Network access (2)
⚙️ External commands (1)
Quality Score
What You Can Build
Build docking libraries
Query ZINC to obtain drug-like compounds, download 3D structures, and prepare libraries for DOCK6 or AutoDock Vina screening.
Find analog compounds
Search by SMILES similarity to discover purchasable analogs of known active compounds for structure-activity relationship studies.
Virtual screening setup
Generate random compound sets filtered by molecular weight, LogP, or H-bond donors for initial virtual screening campaigns.
Try These Prompts
Search the ZINC database for compound ZINC000000000001 and show its SMILES, supplier codes, and catalogs.
Find compounds similar to Ibuprofen (CC(C)Cc1ccc(cc1)C(C)C(=O)O) using a Tanimoto distance of 3. Return ZINC IDs, SMILES, and available catalogs.
Generate a list of 1000 random drug-like compounds from ZINC22 with MW 350-500. Show ZINC IDs, SMILES, and tranche codes.
Find 50 lead-like analogs (MW 250-350) of a hit compound with SMILES CCCc1ccc(cc1)C(C)N. Return purchasable options with supplier codes.
Best Practices
- Use exact SMILES matches first, then expand to similarity searches with distance parameters
- Limit output fields to only needed data to reduce response size and improve performance
- Cache frequently accessed compound data locally to reduce repeated API calls
Avoid
- Making rapid consecutive requests without delays - respect server rate limits
- Requesting all output fields when only a few are needed
- Skipping tranche property parsing - these encode MW, LogP, and H-bond donor counts useful for filtering
Frequently Asked Questions
What is ZINC database?
Do I need an API key?
What file formats are available?
How do I download 3D structures?
What is a tranche?
Can I use this for commercial research?
Developer Details
Author
K-Dense-AILicense
MIT
Repository
https://github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/zinc-databaseRef
main
File structure