openadmet-models contains implementations of machine learning architectures and training routines for use in the OpenADMET project. Our goal is to provide a consistent framework for rapid development, experimentation, and prototyping of ML models for ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) prediction tasks.
The library includes traditional machine learning methods, deep learning models, and active learning workflows. It is designed for general-purpose use and is not intended to implement every state-of-the-art architectures, but rather to provide a practical, flexible foundation for ADMET modeling.
Read the documentation here to learn more. There is also a set of demonstration tutorials for a deeper dive into and a showcase example you can try live on Google Colab!
This library is made available under the MIT open source license.
The development version of openadmet-models can be installed directly from the main branch of this repository.
First install the package dependencies using mamba:
mamba env create -f devtools/conda-envs/openadmet-models.yaml
# or if you want a GPU compatible version devtools/conda-envs/openadmet-models-gpu.yamlThe openadmet-models library can then be installed via:
python -m pip install -e --no-deps .
The OpenADMET development team.
Copyright (c) 2025, OpenADMET Models Contributors
OpenADMET is an Open Molecular Software Foundation hosted project.