Skills chembl-database

🔬

chembl-database

Name: chembl-database
Author: K-Dense-AI

Safe ⚙️ External commands🌐 Network access

Query ChEMBL bioactive molecules and drug discovery data

Also available from: davila7

Access the ChEMBL database to search compounds by structure and properties. Retrieve bioactivity measurements, find inhibitors, and perform structure-activity relationship studies for drug discovery research.

Supports: Claude Codex Code(CC)

🥉 75 Bronze

Download the skill ZIP

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

Toggle on and start using

Test it

Using "chembl-database". Get information about aspirin from ChEMBL

Expected outcome:

ChEMBL ID: CHEMBL25
Preferred Name: ASPIRIN
Molecular Weight: 180.16 g/mol
LogP: 1.19
Drug Type: Small molecule
First Approved: 1930

Using "chembl-database". Find kinase inhibitors with IC50 below 100 nM

Expected outcome:

Retrieved 247 kinase inhibitors with IC50 < 100 nM
Top hits include approved drugs like Imatinib, Gefitinib, and Erlotinib
Results sorted by potency (lowest IC50 first)

Using "chembl-database". Find compounds similar to caffeine

Expected outcome:

Found 156 compounds similar to caffeine (>80% similarity)
Most similar: Theobromine (96%), Theophylline (94%)
Includes natural xanthine derivatives and synthetic analogs

Security Audit

Safe

v4 • 1/17/2026

All 198 static findings are FALSE POSITIVES. The analyzer misidentified markdown code formatting (backticks) as shell command execution, SMILES chemical notation as shell operators, and documentation URLs as network endpoints. This is a legitimate scientific database integration skill containing only documentation for the ChEMBL Web Resource Client library. The Python example file contains only wrapper functions that call the official chembl_webresource_client library. No executable malicious code or dangerous functionality exists.

Files scanned

3,084

Lines analyzed

findings

Total audits

Risk Factors

⚙️ External commands (155)

🌐 Network access (10)

references/api_reference.md:269 references/api_reference.md:270 references/api_reference.md:271 references/api_reference.md:272 skill-report.json:6 SKILL.md:382 SKILL.md:383 SKILL.md:384 SKILL.md:385 SKILL.md:386

Audited by: claude View Audit History →

Quality Score

Architecture

100

Maintainability

Content

Community

100

Security

Spec Compliance

What You Can Build

Find Inhibitors for Drug Targets

Search for potent inhibitors of specific targets like EGFR or kinases by querying bioactivity measurements.

Analyze Drug-Target Interactions

Retrieve drug mechanisms, indications, and activity profiles to understand compound-target relationships.

Structure-Activity Relationship Studies

Find similar compounds and analyze molecular properties to identify structure-activity patterns.

Try These Prompts

Find Molecule by ID

Get information about molecule CHEMBL25 (aspirin) from ChEMBL

Search by Properties

Find all molecules with molecular weight under 500, LogP under 5, and at most 10 hydrogen bond acceptors

Query Bioactivity

Find all EGFR inhibitors with IC50 values below 100 nM

Structure Search

Find compounds similar to aspirin (CC(=O)Oc1ccccc1C(=O)O) with 85 percent similarity

Best Practices

Use caching to respect rate limits when making repeated queries
Apply molecular weight and LogP filters early to reduce result sets
Always check the data_validity_comment field for potential issues

Avoid

Making rapid consecutive requests without respecting rate limits
Ignoring the pchembl_value normalization when comparing activities
Assuming all compounds in ChEMBL are experimentally verified

Frequently Asked Questions

What is ChEMBL?

ChEMBL is a manually curated database of bioactive molecules maintained by EBI, containing over 2 million compounds.

Does this skill require API keys?

No, ChEMBL is a free public database. No authentication is required for basic queries.

What activity values are available?

ChEMBL provides IC50, Ki, Kd, EC50, and other bioactivity measurements with standardized units.

Can I search by chemical structure?

Yes, use SMILES strings for similarity searches and substructure matching.

How are results returned?

Results are lazy-evaluated iterators. Convert to list() to force execution.

What are good property filters for drug-likeness?

MW under 500, LogP under 5, HBA under 10, HBD under 5 follow Lipinski's rule of 5.

Developer Details

Author

K-Dense-AI

License

MIT

Repository

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

Ref

main

File structure

📁 references/

📄 api_reference.md

📁 scripts/

📄 example_queries.py

📄 evaluation_output.json

📄 SKILL.md