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

Predicting protein secondary structure by cascade-correlation neural networks.

Matthew J Wood1, Jonathan D Hirst

  • 1School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UK.

Bioinformatics (Oxford, England)
|February 13, 2004
PubMed
Summary

Cascade-correlation neural networks offer a faster alternative to back-propagation for protein secondary structure prediction. This constructive algorithm achieves comparable accuracy in less training time.

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

Balancing optimism versus potential risks of AI-driven drug discovery.

Expert opinion on drug discovery·2026
Same author

PyMolGen: Database-Driven Molecular Generation of Drug-Like Compounds.

Journal of chemical information and modeling·2026
Same author

Proposed Biosynthesis of the Complex Ring-Fused Diterpene Rameswaralide. Mechanistic Insights Using Density Functional Theory.

The Journal of organic chemistry·2026
Same author

Rega: A Platform for the Prediction of the Regioselectivity of C-H Functionalization Reactions.

Journal of chemical information and modeling·2026
Same author

Revision Outcomes of Mixed Versus Matched Components in Primary Total Hip Arthroplasty: A Population-Based Study.

The Journal of arthroplasty·2026
Same author

Hydrofluoric Acid Cutaneous Burns: A Systematic Review of Emergency Management and General Surgical Sequelae.

Cureus·2026

Area of Science:

  • Computational biology
  • Bioinformatics
  • Machine learning in structural biology

Background:

  • Back-propagation neural networks are widely used for protein secondary structure prediction.
  • Back-propagation can be computationally intensive and slow to train.
  • Efficient prediction of protein secondary structure is crucial for understanding protein function.

Purpose of the Study:

  • To compare the performance of cascade-correlation neural networks against back-propagation for protein secondary structure prediction.
  • To evaluate the speed and accuracy of the cascade-correlation architecture.
  • To identify potentially faster and equally effective machine learning methods for structural bioinformatics.

Main Methods:

  • Utilized the cascade-correlation neural network architecture, a constructive learning algorithm.

Related Experiment Videos

  • Compared predictive accuracy against the established back-propagation neural network algorithm.
  • Measured training time efficiency for both methods.

    • Main Results:

    • Cascade-correlation achieved predictive accuracies comparable to back-propagation.
    • The cascade-correlation architecture demonstrated significantly reduced training times.
    • This suggests a more efficient approach to protein secondary structure prediction.

    Conclusions:

    • Cascade-correlation presents a viable and faster alternative to back-propagation for protein secondary structure prediction.
    • The constructive nature of cascade-correlation offers advantages in learning speed.
    • This research highlights the potential of alternative neural network architectures in bioinformatics.