ProTox

ProTox predicts rodent oral toxicity of small molecules by estimating LD50 through compound similarity and toxicophore-based pharmacophore models.


Key Features:

  • Prediction methodology: Estimates median lethal dose (LD50) by analyzing chemical similarity to compounds with known LD50 and by identifying toxic fragments within two-dimensional chemical structures.
  • Toxicophore pharmacophores: Uses an in-house collection of protein-ligand-based pharmacophore models termed "toxicophores" that are associated with targets linked to adverse drug reactions.
  • Input data: Operates on submitted two-dimensional representations of compound structures.
  • Performance evaluation: Validated on an external dataset with reported sensitivity 76%, specificity 95%, and precision 75%.

Scientific Applications:

  • Preclinical toxicity prediction: Predicts rodent oral toxicity to support hazard assessment of small-molecule drug candidates and chemicals.
  • Mechanistic insight and target association: Identifies toxic fragments and toxicophore-associated targets to inform mechanistic hypotheses and follow-up experimental investigations.

Methodology:

Computational steps include compound similarity analysis to known LD50 values, toxic fragment identification within 2D structures, and application of protein-ligand-based pharmacophore ("toxicophore") models; performance was assessed using an external validation set reporting sensitivity, specificity, and precision metrics.

Topics

Details

Tool Type:
web application
Operating Systems:
Linux, Windows, Mac
Programming Languages:
Python
Added:
5/16/2017
Last Updated:
12/10/2018

Operations

Publications

Drwal MN, Banerjee P, Dunkel M, Wettig MR, Preissner R. ProTox: a web server for the in silico prediction of rodent oral toxicity. Nucleic Acids Research. 2014;42(W1):W53-W58. doi:10.1093/nar/gku401. PMID:24838562. PMCID:PMC4086068.

Documentation