πŸ—‚οΈ Navigation

ADMET-AI

A machine learning ADMET platform for evaluation of large-scale chemical libraries.

Visit Website β†’

Overview

ADMET-AI is a platform for ADMET prediction that offers both an easy-to-use website and a powerful Python package. It uses graph neural networks to predict a wide range of ADMET properties and is designed for high-throughput screening of large chemical libraries. It also provides context for its predictions by comparing them to the properties of approved drugs.

✨ Key Features

  • Prediction of 41 ADMET properties
  • Fast prediction speeds
  • Web interface for up to 1,000 molecules
  • Python package for large-scale prediction
  • Comparison of predictions to approved drugs from DrugBank
  • Open-source

🎯 Key Differentiators

  • High prediction speed
  • Availability as a Python package for large-scale use
  • Use of modern graph neural network architecture
  • Contextualization of predictions with DrugBank data

Unique Value: Provides a free, fast, and accurate open-source platform for large-scale ADMET prediction, bridging the gap between web-based tools and the needs of computational drug discovery.

🎯 Use Cases (4)

High-throughput virtual screening Generative AI in drug discovery Academic research Filtering of large chemical libraries

βœ… Best For

  • Models have shown high performance on the TDC ADMET Benchmark Group leaderboard.

πŸ’‘ Check With Vendor

Verify these considerations match your specific requirements:

  • Not suitable for use in regulated environments.
  • Web interface has a limit on the number of molecules.

πŸ† Alternatives

SwissADME ADMETlab 2.0 pkCSM

Offers superior speed and scalability for large datasets compared to other free web servers.

πŸ’» Platforms

Web Python package

βœ… Offline Mode Available

πŸ”Œ Integrations

Python

πŸ’° Pricing

Contact for pricing
Free Tier Available

Free tier: Web interface limited to 1,000 molecules per prediction.

Visit ADMET-AI Website β†’