skills.rest
mims-harvard
mims-harvardOFFICIAL
Explore all skills by this author →
Website𝕏@marinkazitnikGitHub

tooluniverse-admet-prediction

Profiles ADMET properties and toxicity of drug candidates from SMILES or compound names.

1.7k|254|Updated Mar 3, 2025
Search Tags:#drug discovery#drug-likeness#smiles#PubChem#admet#pharmacokinetics#toxicity prediction
One-click install
npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-admet-prediction
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tooluniverse-admet-prediction
Source: https://github.com/mims-harvard/ToolUniverse/tree/main/plugins/tooluniverse/skills/tooluniverse-admet-prediction
Command: npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-admet-prediction

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires admet-ai.

What problem does it solve?

Evaluating whether a drug candidate can be absorbed, distributed, metabolized, excreted, and tolerated safely requires querying many separate prediction models and toxicity databases, then reconciling conflicting evidence into a single decision.

Core Features & Use Cases

  • Five-phase ADMET profiling: resolves compound identity via PubChem, then runs physicochemical, ADME, toxicity, and clinical-context analysis ending in a 13-category pass/warn/fail scorecard.
  • Multi-source evidence integration: combines ADMET-AI predictions, SwissADME drug-likeness rules, PubChemTox experimental data, and ChEMBL clinical phase data with T1-T4 evidence grading.
  • Use Case: Given a SMILES string for a screening hit, produce a full pharmacokinetic and toxicity report covering BBB penetration, CYP interactions, hERG liability, AMES mutagenicity, and Lipinski compliance before committing to lab testing.

Quick Start

Ask the agent to run a full ADMET profile and drug-likeness scorecard for the compound named ibuprofen.

Frequently Asked Questions about tooluniverse-admet-prediction

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

FAQPage Schema
How do I predict ADMET properties of a compound from a SMILES string?▼

Provide the SMILES string and the workflow resolves its PubChem CID, then runs ADMET-AI predictions for physicochemical properties, BBB penetration, bioavailability, CYP interactions, and toxicity endpoints, ending in a pass/warn/fail scorecard.

What tools does this ADMET profiling workflow combine?▼

It integrates ADMET-AI machine learning predictions, SwissADME drug-likeness rules such as Lipinski and Veber, PubChemTox experimental toxicity data, and ChEMBL clinical phase information, each tagged with an evidence tier from T1 to T4.

Why do ADMET-AI tools fail with an admet-ai package error?▼

ADMET-AI tools require the optional ml extra installed via uv pip install 'tooluniverse[ml]'. Without it the tools appear in the tool list but fail at call time; run tooluniverse-doctor to confirm which optional groups are installed.

Can I assess drug-likeness without the ADMET-AI models installed?▼

Yes. SwissADME provides Lipinski, Veber, Ghose, Egan, and Muegge rule compliance plus PAINS alerts independently, and PubChemTox supplies experimental toxicity data, so the workflow falls back to these sources when ADMET-AI is unavailable.

What toxicity endpoints are checked for a drug candidate?▼

The workflow reports AMES mutagenicity, DILI hepatotoxicity, hERG cardiotoxicity, clinical toxicity, predicted LD50, carcinogenicity, skin sensitization, nuclear receptor activity, and stress response pathways, cross-checked against experimental PubChemTox data.

Related Skills

Explore More Education & Research Skills→
learningmatter-mitlearningmatter-mit
144

chem-similarity-search

Query PubChem's fastsimilarity_2d engine for structurally similar compounds from SMILES or CID.

Official
Basic
robotlearning123robotlearning123

medchem

Filter compound libraries using medicinal chemistry rules and structural alerts.

Community
Advanced
Rawgrowth-ConsultingRawgrowth-Consulting

drug-discovery

Consolidates medicinal chemistry data into a drug-discovery briefing using ChEMBL, Lipinski/Veber metrics, OpenFDA, ADMET interpretation, and OpenTargets disease associations.

Official
Advanced
InternScienceInternScience
52

adme-prediction

Predict ADME properties from SMILES strings using Morgan fingerprints and Random Forest models.

Official
Advanced