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

Neural network classification of mutagens using structural fragment data

M Brinn1, P T Walsh, M P Payne

  • 1Research and Laboratory Services Division, Health and Safety Executive, Sheffield S3 7HQ, UK

SAR and QSAR in Environmental Research
|January 1, 1993
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

IL-36α expression is elevated in ulcerative colitis and promotes colonic inflammation.

Mucosal immunology·2016
Same author

Prediction of acute aquatic toxicity in Tetrahymena pyriformis--'Eco-Derek', a knowledge-based system approach.

SAR and QSAR in environmental research·2013
Same author

Survey of the reliability of carbon monoxide alarms deployed in domestic homes and efficacy of use by consumers.

Indoor air·2012
Same author

Evaluation of portable single-gas monitors for the detection of low levels of hydrogen sulfide and sulfur dioxide in petroleum industry environments.

Journal of occupational and environmental hygiene·2012
Same author

Response of electrochemical oxygen sensors to inert gas-air and carbon dioxide-air mixtures: measurements and mathematical modelling.

Journal of hazardous materials·2010
Same author

Development of an expert system rulebase for identifying contact allergens.

Toxicology in vitro : an international journal published in association with BIBRA·2010

A neural network model was developed to predict mutagenicity by analyzing chemical structures. The system effectively identified structural patterns, classifying mutagens and non-mutagens with high accuracy.

Area of Science:

  • Computational chemistry
  • Toxicology
  • Machine learning

Background:

  • Understanding structure-property relationships is crucial for predicting chemical toxicity.
  • Identifying mutagens is essential for public health and drug development.

Purpose of the Study:

  • To investigate structure-property relationships using a neural network approach.
  • To develop a predictive model for mutagenicity based on chemical substructures.

Main Methods:

  • A neural network, specifically back-propagation networks, was employed.
  • A dataset of 1280 substructural fragments from mutagens and non-mutagens was used as input.
  • An algorithm optimized fragment selection to enhance predictive power.

Main Results:

Related Experiment Videos

  • The developed neural network achieved approximately 11% failure rate on the test set and 6% on the training set.
  • Analysis of the hidden layer revealed clusters of mutagens and non-mutagens.
  • Some clusters corresponded to known mutagenic and non-mutagenic structural classes.

Conclusions:

  • The neural network model demonstrated effectiveness in classifying mutagens and non-mutagens.
  • The approach provides insights into how the network identifies mutagenic structural features.
  • This method offers a valuable tool for predicting chemical mutagenicity.