Can I filter compounds by drug-likeness?

Yes, you can filter results using the 'tranche' system which organizes compounds by H-bond donors, LogP, and molecular weight, or use subsets like 'lead-like' and 'drug-like'.

What is the ZINC Database skill for Claude Code?

It is a specialized capability that allows Claude to programmatically query the ZINC database, a repository of over 230 million purchasable compounds for scientific research.

Can I perform similarity searches with this skill?

Yes, the skill supports similarity and analog searches using SMILES notation and Tanimoto distance parameters to find structurally related molecules.

Which version of ZINC does this skill use?

This skill primarily focuses on ZINC22, the most current and comprehensive version of the database maintained by UCSF.

Does this skill provide 3D structures for molecular docking?

While the API primarily returns metadata and SMILES, the skill provides the necessary paths and identifiers to retrieve 3D-ready MOL2 and SDF files from ZINC's file repositories.

ZINC Scientific Database

Name: ZINC Scientific Database
Author: lifangda

bylifangda

•

Data Science & ML

Accesses the ZINC database of 230M+ purchasable compounds for drug discovery, virtual screening, and molecular analog searching.

About

This skill integrates the UCSF ZINC database directly into Claude, enabling researchers and developers to perform sophisticated chemical searches and data retrieval tasks. It provides automated access to the ZINC22 repository, allowing for SMILES-based similarity searches, batch ZINC ID lookups, and the identification of commercially available analogs. Whether you are preparing molecular docking libraries, exploring chemical space via random sampling, or retrieving 3D-ready structures for virtual screening, this skill streamlines the interface between AI-assisted coding and computational chemistry workflows.

Key Features

Perform Tanimoto-based similarity searches to identify chemical analogs
Search over 230 million compounds by ZINC ID, SMILES, or supplier codes
Sample chemical space using pre-defined lead-like, drug-like, or fragment subsets
Retrieve molecular property 'tranches' including LogP, MW, and H-bond donors
Access direct links to 3D-ready structures in MOL2 and SDF formats for docking
16 GitHub stars

Use Cases

Identifying commercially available analogs for lead optimization in drug discovery
Automating the retrieval of chemical metadata and supplier information for laboratory purchasing
Building and filtering compound libraries for high-throughput virtual screening

About

Key Features

Perform Tanimoto-based similarity searches to identify chemical analogs
Search over 230 million compounds by ZINC ID, SMILES, or supplier codes
Sample chemical space using pre-defined lead-like, drug-like, or fragment subsets
Retrieve molecular property 'tranches' including LogP, MW, and H-bond donors
Access direct links to 3D-ready structures in MOL2 and SDF formats for docking
16 GitHub stars

Use Cases

Identifying commercially available analogs for lead optimization in drug discovery
Automating the retrieval of chemical metadata and supplier information for laboratory purchasing
Building and filtering compound libraries for high-throughput virtual screening