PROTOX - Prediction of Rodent Oral TOXicity


Welcome to PROTOX, a webserver for the prediction of oral toxicities of small molecules in rodents!
The prediction of compound toxicities is an important part of the drug design development process. Computational toxicity estimations are not only faster than the determination of toxic doses in animals, but can also help to reduce the amount of animal experiments. To read more about reducing animal testing, go to Animal Ethics 3R.
Our webserver is based on chemical similarities between compounds with known toxic effects and the presence of toxic fragments. To predict the oral toxicity of your compounds, please click here. For a description of the server and tutorials, go to FAQ. To see statistics about our training set as well as the cross-validation results, please go to Statistics.
Should you have further questions, do not hesitate to contact us!


Drwal M.N., Banerjee P., Dunkel M., Wettig M.R., Preissner R.: ProTox: a web server for the in silico prediction of rodent oral toxicity
Nucleic Acids Res (Web server issue 2014); NAR

Schmidt U., Struck S., Gruening B., Hossbach J., Jaeger I.S., Parol R., Lindequist U., Teuscher E., Preissner R.: SuperToxic: a comprehensive database of toxic compounds.
Nucleic Acids Res 37 (Database issue 2009): D295-9; NAR

Predict compound toxicity


PROTOX enables uncomplicated predictions of oral rodent toxicities. To conduct a toxicity prediction, please click on the image.

Toxic doses and toxicity classes

Toxic doses are often given as LD50 values in mg/kg body weight. The LD50 is the median lethal dose meaning the dose at which 50% of test subjects die upon exposure to a compound.

Toxicity classes are defined according to the globally harmonized system of classification of labelling of chemicals (GHS):
  • Class I: fatal if swallowed (LD50 ≤ 5 mg/kg)
  • Class II: fatal if swallowed (5 < LD50 ≤ 50 mg/kg)
  • Class III: toxic if swallowed (50 < LD50 ≤ 300 mg/kg)
  • Class IV: harmful if swallowed (300 < LD50 ≤ 2000 mg/kg)
  • Class V: may be harmful if swallowed (2000 < LD50 ≤ 5000 mg/kg)
  • Class VI: non-toxic (LD50 > 5000 mg/kg)

Toxicity targets

Toxicity targets are protein targets which have been associated with adverse drug reactions and toxic effects. Here, we predict possible binding to toxicity targets using a collection of protein-ligand-based pharmacophores. To see warning about possible toxicity targets of your input compound, please click on Tox Prediction.


- September 2016: Our paper on toxicity prediction has been accepted- Computational methods for prediction of in vitro effects of new chemical structures Journal of Chemoinformatics
- September 2015: To download an example KNIME workflow we used in the Tox21 challenge, please go to Downloads.
- July 2015: Our paper about the Tox21 challenge has been accepted in a special issue of Frontiers in Environmental Science
- January 2015: We have participated in the Tox21 Data Challenge 2014 and our prediction models show the best balanced accuracy for 4 targets!
- September 2014: New tox target has been added: Opioid receptor mu-type - the target has been homology modelled using a structure of the mouse protein
- August 2014: Due to new X-ray structures, new tox target pharmacophores added: Androgen receptor (1), Estrogen receptor 1 (3)
- July 2014: ProTox paper published in NAR web server issue