Absorption, Distribution, Metabolism, Excretion and Toxicity (ADMET) properties of drug candidates or environmental chemicals play a key role in drug discovery and environmental hazard assessment. The ADMET structure-activity relationship server, entitled admetSAR, is a comprehensive knowledge and tool for predicting ADMET properties of drug candidates and environmental chemicals. In our server, over 200,000 ADMET annotated data points for about 96 thousand of unique compounds have been manually curated from large literatures. The admetSAR server provides a user-friendly interface to easily search for chemical profiles, by CASRN, common name and similarity search.

The new version of admetSAR (version 2.0) mainly focuses on in silico prediction of chemical ADMET properties. More than 40 predictive models were implemented in admetSAR for new chemical ADMET properties in silico filtering. These models are trianed by state-of-the-art machine learning methods including support vector machine, k-nearest neighbors, nueral network, etc.

The tutorial can be downloaded here. Contribution to the tutorial is welcome. Please click here to see the source. For feedback, suggestions, or for bug reports, feel free to contact us.

admetSAR (v2) was developed by Hongbin Yang (current maintainer) with the contributions from many people including Chaofeng Lou, Jie Li, Chen Zhang, Defang Fan, Feixiong Cheng, Fuxing Li, Lixia Sun, Qianqian Cao, Yadi Zhou, Zengrui Wu, Zhuang Wang.
admetSAR was originally created by Feixiong Cheng and Yadi Zhou.
The project is proposed and supervised by Prof. Yun Tang.