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

Data modelling with neural networks: advantages and limitations

D J Livingstone1, D T Manallack, I V Tetko

  • 1Centre for Molecular Design, University of Portsmouth, U.K.

Journal of Computer-Aided Molecular Design
|March 1, 1997
PubMed
Summary

Artificial neural networks (ANNs) offer powerful nonlinear data modeling for drug design, addressing challenges like overfitting and enabling variable selection. This review details their operation, applications, and methods to overcome common data modeling problems.

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

QSAR-derived affinity fingerprints (part 1): fingerprint construction and modeling performance for similarity searching, bioactivity classification and scaffold hopping.

Journal of cheminformatics·2021
Same author

IRAK3 modulates downstream innate immune signalling through its guanylate cyclase activity.

Scientific reports·2019
Same author

The acid/base profile of agrochemicals.

SAR and QSAR in environmental research·2017
Same author

Use of Kv1.3 blockers for inflammatory skin conditions.

Current medicinal chemistry·2010
Same author

Computational techniques for the prediction of toxicity.

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

Synthesis and preliminary screening of novel tryptamines as 5-HT4 receptor ligands.

Current medicinal chemistry·2010

Area of Science:

  • Computational chemistry
  • Cheminformatics
  • Artificial intelligence

Background:

  • Artificial neural networks (ANNs) have historical roots and specific operational principles.
  • Early applications of ANNs in data modeling for drug design are documented.

Purpose of the Study:

  • To review the origins and operation of ANNs.
  • To discuss their application in data modeling for drug design.
  • To address common problems encountered when using ANNs for data modeling.

Main Methods:

  • Description of ANN origins and operation.
  • Review of early applications in drug design data modeling.
  • Discussion of overfitting, chance effects, overtraining, and interpretation issues.
  • Illustrative examples of avoiding overfitting, chance effects, and overtraining.

Related Experiment Videos

  • Demonstration of ANNs as variable selection tools.
  • Exploration of ANNs for nonlinear data modeling.
  • Visualization of multivariate data in two dimensions using ANNs.

    • Main Results:

    • Methods for avoiding overfitting, chance effects, and overtraining in ANN models are presented.
    • The utility of ANNs as a variable selection tool is demonstrated.
    • The advantage of ANNs in nonlinear data modeling is highlighted.
    • Multivariate data from charge transfer complexes were successfully displayed in two dimensions using ANNs.

    Conclusions:

    • ANNs are effective nonlinear data modeling tools with applications in drug design.
    • Strategies exist to mitigate common challenges like overfitting and overtraining.
    • ANNs can simplify complex datasets for better interpretation and visualization.