Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Modeling toxicity by using supervised kohonen neural networks.

Paolo Mazzatorta1, Marjan Vracko, Aneta Jezierska

  • 1Istituto Mario Negri, via Eritrea 62, 20157 Milan, Italy. mazzatorta@marionegri.it

Journal of Chemical Information and Computer Sciences
|March 26, 2003
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Scientific Opinion on Benzophenone - 4 (CAS No. 4065-45-6, EC No. 223-772-2) used in cosmetics products - SCCS/1660/23.

NAM journal·2026
Same author

SCCS opinion on biphenyl-2-ol and sodium 2-biphenylolate used in cosmetic products (CAS/EC No. 90-43-7/201-993-5 and 132-27-4/205-055-6)- SCCS/1669/24.

NAM journal·2026
Same author

In silico prediction of Ames mutagenicity for organosilicon compounds: Exploring and enhancing chemical space boundaries.

Regulatory toxicology and pharmacology : RTP·2026
Same author

Simulation of fish chronic toxicity using the Las Vegas algorithm and the vector of ideality of correlation.

Environmental toxicology and chemistry·2026
Same author

An in silico study of binding of new lutein esters to nuclear endocrine receptors.

Food research international (Ottawa, Ont.)·2026
Same author

A framework for chemical hazard assessments under 'Safe and Sustainable by Design' using multiple in silico tools.

Integrated environmental assessment and management·2026
Same journal

Future Papers.

Journal of chemical information and computer sciences·2016
Same journal

Future Papers.

Journal of chemical information and computer sciences·2016
Same journal

Future Papers.

Journal of chemical information and computer sciences·2016
Same journal

Future Papers.

Journal of chemical information and computer sciences·2016
Same journal

Future Papers.

Journal of chemical information and computer sciences·2016
Same journal

Future Papers.

Journal of chemical information and computer sciences·2016
See all related articles

Counterpropagation neural networks effectively predict chemical toxicity. This computational toxicology approach uses a large dataset to identify potential hazards and visualize chemical data distributions.

Area of Science:

  • Computational toxicology
  • Cheminformatics
  • Machine learning in drug discovery

Background:

  • Assessing chemical toxicity is crucial for drug development and environmental safety.
  • Traditional toxicity testing methods can be time-consuming and expensive.
  • Predictive models offer a promising alternative for early-stage toxicity screening.

Purpose of the Study:

  • To evaluate the efficacy of counterpropagation neural networks (CPNNs) for predicting chemical toxicity.
  • To develop and validate a CPNN model using a curated dataset of chemical compounds.
  • To explore the utility of CPNNs in data visualization for toxicological analysis.

Main Methods:

  • A dataset of 568 chemicals was curated for toxicity investigation.
  • A counterpropagation neural network model was trained on 282 chemicals and tested on 286.

Related Experiment Videos

  • The model's performance was evaluated using the R-squared (R²) metric.

    • Main Results:

    • The developed CPNN model achieved a high overall performance with R² = 0.83.
    • The model demonstrated strong predictive accuracy on the training set (R² = 0.97).
    • The model showed moderate performance on the unseen test set (R² = 0.59), indicating generalization capability.
    • The technique generated 2D maps for data distribution, clustering, and outlier identification.

    Conclusions:

    • Counterpropagation neural networks are a powerful tool for investigating chemical toxicity.
    • CPNNs provide valuable insights into chemical data distribution and facilitate outlier identification.
    • This approach supports computational toxicology and enhances the efficiency of toxicity assessment.