medchem

Apply medicinal chemistry rules and structural alerts to triage drug-like compounds.

15|2|Updated Dec 17, 2025
One-click install
npx skills add https://github.com/rubensliv/k-dense-ai --skill medchem-rubensliv
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: medchem
Source: https://github.com/rubensliv/k-dense-ai/tree/main/scientific-skills/medchem
Command: npx skills add https://github.com/rubensliv/k-dense-ai --skill medchem-rubensliv

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pandas, datamol, medchem, rdkit, tqdm, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Medicinal chemistry screening involves applying drug-likeness rules, PAINS filters, structural alerts, and complexity metrics to triage large compound libraries. This Skill helps researchers quickly identify promising candidates and discard problematic structures, accelerating lead discovery workflows.

Core Features & Use Cases

  • Medicinal chemistry rules: Apply established Ro5/Ro3/Oprea/CNS/Veber/lead-like rules to assess drug-likeness.
  • Structural alerts & PAINS: Detect potentially problematic patterns with CommonAlertsFilters, NIBR, and Lilly demerits.
  • Complexity & constraints: Compute synthetic complexity and apply property constraints to prioritize compounds.
  • High-level APIs: Use RuleFilters, functional modules, and query language to compose custom filtering pipelines.
  • Use Case: Screen a library of 10,000 molecules to select CNS candidates with acceptable complexity and no common alerts.

Quick Start

Filter a list of molecules through medchem rules, alerts, and complexity metrics to obtain a clean, prioritized candidate set.

Frequently Asked Questions about medchem

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I apply PAINS filters and structural alerts to a large molecule library?

Apply PAINS filters and structural alerts to molecule libraries using this Skill's CommonAlertsFilters and NIBR modules to detect problematic patterns. It supports high-throughput filtering on large compound sets to discard problematic structures during lead discovery.

What medicinal chemistry rules can I use to evaluate drug-likeness for lead optimization?

Evaluate drug-likeness using built-in Ro5, Ro3, Oprea, CNS, Veber, and lead-like rules. These medicinal chemistry rules assess compound viability to help prioritize candidates for lead optimization workflows.

Can I use RDKit and datamol for high-throughput compound screening?

Yes, this Skill requires RDKit and the medchem library, while optionally using datamol and pandas for data handling. It enables high-throughput compound screening with parallel processing to triage large libraries efficiently.

How do I filter molecules for CNS candidates with acceptable synthetic complexity?

Filter CNS candidates by applying CNS-focused screening rules alongside synthetic complexity metrics. This Skill computes complexity scores and applies property constraints to select suitable compounds from libraries of thousands of molecules.

Does this Skill support custom filtering pipelines for structural alerts?

Yes, custom filtering pipelines are supported through RuleFilters, functional modules, and a query language. You can compose tailored pipelines to apply specific medicinal chemistry rules and structural alerts to your compounds.

What's the best way to triage 10,000 molecules for lead-like properties?

Triage large molecule libraries by applying lead-like rules, structural alerts, and complexity metrics in parallel. This Skill processes 10,000 molecules to select clean, prioritized candidate sets for lead discovery.