What are the primary rules supported by the Medchem skill?

It supports industry-standard criteria including Lipinski's Rule of Five, Veber, Oprea, CNS, lead-like (soft/strict), Rule of Three, and Reos.

Can I use Medchem for large molecular datasets?

Yes, the skill supports parallel processing using all available CPU cores, making it highly efficient for triaging libraries containing thousands of molecules.

Can I create custom filtering logic?

Absolutely. Medchem features a specialized query language and a constraints module to define specific Molecular Weight, LogP, TPSA, and other property bounds.

Does it detect reactive functional groups?

Yes, it includes specialized modules for PAINS, NIBR structural alerts, and Lilly Demerits to identify potentially toxic, reactive, or problematic structural motifs.

What molecular formats are compatible with this skill?

The skill primarily works with SMILES strings and RDKit-compatible molecular objects through the datamol library integration.

MedChem Drug Discovery Filters

Name: MedChem Drug Discovery Filters
Author: Microck

byMicrock

•

Data Science & ML

Filters and prioritizes molecular compound libraries using established medicinal chemistry rules and structural alerts.

Medchem is a specialized scientific skill designed for drug discovery and cheminformatics workflows, enabling users to evaluate chemical compounds against industry-standard criteria. It provides an automated way to apply drug-likeness rules such as Lipinski’s Rule of Five, detect problematic structural patterns like PAINS, and calculate molecular complexity metrics. By integrating these medicinal chemistry filters directly into Claude's environment, researchers can efficiently triage large compound datasets, identify high-quality leads, and ensure that chemical candidates meet specific pharmacokinetic and safety guidelines before proceeding to synthesis or bioassays.

Key Features

01Apply classic drug-likeness rules including Lipinski, Veber, and Oprea

02Filter structural alerts and PAINS patterns automatically

03Calculate molecular complexity and synthetic accessibility metrics

04Identify specific functional groups and hinge/phosphate binders

05Execute complex filtering logic using the Medchem Query Language

0681 GitHub stars

Use Cases

01Initial triage of large virtual compound libraries for drug-like candidates

02Lead optimization to ensure candidates maintain favorable ADME profiles

03Identifying and removing reactive or toxic structural motifs from screening sets

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add microck/ordinary-claude-skills medchem

For use in Claude.ai and ChatGPT

Download Skill