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

Simple consensus procedures are effective and sufficient in secondary structure prediction.

Mario Albrecht1, Silvio C E Tosatto, Thomas Lengauer

  • 1Max-Planck-Institute for Informatics, Stuhlsatzenhausweg 85, 66123 Saarbrücken, Germany. mario.albrecht@mpi-sb.mpg.de

Protein Engineering
|August 14, 2003
PubMed
Summary

Combining three advanced protein secondary structure prediction methods using majority voting improves prediction accuracy by up to 1.5 percentage points. This consensus approach enhances confidence in predicted structures without significantly impacting performance with filtering.

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

On the state of protein function prediction: a report on the fourth CAFA challenge.

bioRxiv : the preprint server for biology·2026
Same author

Tandem repeats matter for the functional versatility of giant proteins.

Trends in biochemical sciences·2026
Same author

Comparative Multi-Marker Environmental DNA Metabarcoding of Marine Metazoan Communities: Water vs. Sediment.

Molecular ecology resources·2026
Same author

Open and sustainable AI: challenges, opportunities and the road ahead in the life sciences.

Nature methods·2026
Same author

Exploring proteins and protein-ligand complexes through residue interaction networks.

Nature protocols·2026
Same author

Effects of Mutations on Tandem-Repeat Proteins Conformation Mechanisms. Application to the Phosphatase PP2A.

Journal of chemical information and modeling·2026

Area of Science:

  • Computational biology
  • Structural bioinformatics
  • Bioinformatics algorithms

Background:

  • Accurate protein secondary structure prediction is crucial for understanding protein function.
  • State-of-the-art prediction methods offer high accuracy but can be improved through consensus.
  • Combining multiple prediction methods can leverage their individual strengths.

Purpose of the Study:

  • To evaluate the performance of majority voting for consensus secondary structure prediction.
  • To determine the impact of combining three leading prediction methods.
  • To assess the effect of filtering short predicted elements on accuracy.

Main Methods:

  • Analysis of majority voting on minimal combination sets of three secondary structure prediction tools.

Related Experiment Videos

  • Utilized three large benchmark datasets from the EVA server.
  • Applied a filtering procedure for short predicted secondary structure elements.

    • Main Results:

    • Consensus prediction using majority voting significantly improved average Q3 accuracy by up to 1.5 percentage points.
    • The similarity between prediction methods was analyzed, revealing higher confidence in consistently predicted structures.
    • Trivial filtering of short elements did not substantially alter prediction accuracy.

    Conclusions:

    • Majority voting on minimal sets of state-of-the-art methods is an effective strategy for enhancing protein secondary structure prediction accuracy.
    • Consensus prediction increases confidence in the reliability of predicted secondary structure elements.
    • Further improvements in prediction accuracy can be achieved by combining complementary prediction algorithms.