What specific drug-likeness rules does this skill support?

The skill supports industry-standard rules including Lipinski's Rule of Five, Rule of Oprea, CNS-likeness, Rule of Lead-like, Rule of Three, Rule of Veber, and the Golden Triangle.

Does it detect Pan-Assay Interference Compounds (PAINS)?

Yes, it includes specialized PAINS filters and structural alert systems like NIBR filters and Lilly Demerits to identify potentially problematic or reactive molecules.

Can I use this for high-throughput screening of large datasets?

Yes, the skill is designed for efficiency and supports parallelization (using n_jobs=-1) to process large SMILES libraries or SDF files quickly.

What is the Medchem Query Language?

It is a specialized syntax that allows you to combine multiple filtering criteria (e.g., 'rule_of_five AND NOT common_alerts') into a single, human-readable string for complex library triage.

Medicinal Chemistry & Molecular Filtering

Name: Medicinal Chemistry & Molecular Filtering
Author: pur3v4d3r

bypur3v4d3r

•

Data Science & ML

Applies industry-standard medicinal chemistry rules and structural alerts to filter and prioritize molecular compound libraries.

The Medchem skill provides a robust toolkit for drug discovery workflows, enabling developers and researchers to triage large chemical libraries using established drug-likeness rules like Lipinski's Rule of Five, Veber, and CNS criteria. It automates the detection of problematic structural patterns through PAINS filters, NIBR alerts, and Lilly Demerits, while also calculating molecular complexity metrics for synthetic accessibility. By integrating these specialized filters into Claude Code, users can efficiently assess compound quality, identify reactive functional groups, and ensure chemical candidates meet specific lead-optimization criteria.

Key Features

011 GitHub stars

02Lilly Demerit scoring for assessing compound quality and safety

03Molecular complexity metrics to evaluate synthetic accessibility

04Specialized Medchem Query Language for building complex filtering logic

05Advanced structural alert detection including PAINS and NIBR filters

06Comprehensive drug-likeness rule checking (Lipinski, Veber, Ro3, CNS)

Use Cases

01Identifying and filtering out problematic PAINS patterns or reactive functional groups

02Prioritizing lead candidates based on medicinal chemistry constraints and complexity

03Initial triage of large compound libraries to identify drug-like candidates

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add pur3v4d3r/pur3-pkb-codebase medchem

For use in Claude.ai and ChatGPT

Download Skill