Does it support parallel processing for large datasets?

Yes, Medchem is designed for scale and supports parallelization across CPU cores using the n_jobs parameter to handle libraries with thousands of molecules efficiently.

Which drug-likeness rules are supported by this skill?

Medchem supports several established rules including Lipinski's Rule of Five, Veber's rules, Rule of CNS, Rule of Three, Oprea, and both soft and strict Lead-like criteria.

What is the Medchem Query Language?

It is a specialized syntax that allows you to combine multiple filtering criteria using logical operators, such as 'rule_of_five AND NOT common_alerts', for complex prioritization tasks.

Can I use this for PAINS filter screening?

Yes, the skill includes built-in Pan-Assay Interference Compounds (PAINS) filters, along with NIBR filters and the Lilly Demerits system for detecting problematic chemical patterns.

How are molecular complexity scores calculated?

The skill provides metrics such as Bertz, Whitlock, and Barone complexity, which help approximate synthetic accessibility and molecular sophistication.

Medchem Molecular Filters

Name: Medchem Molecular Filters
Author: lifangda

bylifangda

•

数据科学与机器学习

Applies medicinal chemistry rules and structural alerts to triage and prioritize compound libraries for drug discovery workflows.

Medchem is a specialized cheminformatics skill that provides a robust suite of tools for molecular filtering and prioritization based on established medicinal chemistry principles. It enables researchers to apply drug-likeness rules such as Lipinski’s Rule of Five, detect structural alerts including PAINS patterns, and calculate molecular complexity metrics using a high-performance Python framework. By integrating these capabilities into Claude's workflow, users can efficiently triage large compound libraries, assess lead optimization candidates, and detect problematic functional groups using production-ready patterns and a specialized query language.

主要功能

01Molecular complexity and synthetic accessibility metric calculation

02Comprehensive drug-likeness filtering (Lipinski, Veber, Rule of CNS, Lead-like)

03High-performance parallel processing for large-scale compound library triaging

04Advanced structural alert detection including PAINS, NIBR, and Lilly Demerits

05Medchem Query Language for building complex, multi-criteria molecular filters

0616 GitHub stars

使用场景

01Initial triage of large compound libraries to identify drug-like leads

02Automating the detection of reactive functional groups or interfering patterns in chemical data

03Filtering molecules by structural alerts and demerits during lead optimization

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lifangda/claude-plugins medchem

For use in Claude.ai and ChatGPT

Download Skill