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

Prediction of protein subcellular localization.

Chin-Sheng Yu1, Yu-Ching Chen, Chih-Hao Lu

  • 1Department of Biological Science and Technology, National Chiao Tung University, Hsinchu, Taiwan, Republic of China.

Proteins
|June 6, 2006
PubMed
Summary

Predicting protein subcellular localization is crucial for understanding protein function. This study introduces a novel two-level support vector machine (SVM) system that accurately predicts localization, outperforming existing methods and offering a reliable tool for bioinformatics.

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

Author Correction: Plasma membrane curvature regulates the formation of contacts with the endoplasmic reticulum.

Nature cell biology·2026
Same author

Zfand5 terminates TLR3/4 signaling and necroptosis by targeting TRIF to the proteasome for degradation.

Cell death & disease·2026
Same author

Membrane Curvature Activates Src kinase and Promotes Metastatic Cancer Cell Survival.

bioRxiv : the preprint server for biology·2026
Same author

Effects of Sperm DNA Fragmentation on Embryonic Cleavage and Semen Parameters in In Vitro Fertilization Treatment.

Journal of aging research·2026
Same author

FCHo2, not talin, enables inside-out activation of integrin ɑvβ5 in curved adhesions.

Nature communications·2026
Same author

Early Enrollment in Diabetes Pay-for-Performance Program Reduced Loss of Life Expectancy in Newly-Diagnosed Patients with Type 2 Diabetes Mellitus (Diabetes Metab J 2025;49:1051-63).

Diabetes & metabolism journal·2026

Area of Science:

  • Computational Biology and Bioinformatics
  • Molecular and Cellular Biology
  • Machine Learning in Biology

Background:

  • Protein function is intrinsically linked to its subcellular localization.
  • Accurate prediction of subcellular localization from protein sequences aids in inferring protein functions.
  • Existing computational tools show variable success and may be biased by homologous sequences in benchmark datasets.

Purpose of the Study:

  • To develop a novel computational approach for predicting protein subcellular localization directly from amino acid sequences.
  • To address the overestimation of performance caused by highly homologous sequences in current benchmark datasets.
  • To create a robust and accurate tool for predicting subcellular localization, unaffected by sequence homology.

Main Methods:

Related Experiment Videos

  • Development of a two-level support vector machine (SVM) classification system.
  • The first level uses multiple SVM classifiers, each trained on specific sequence-derived feature vectors.
  • The second level employs a jury SVM classifier to integrate predictions and generate probability distributions for localizations.

Main Results:

  • The proposed two-level SVM system demonstrated significantly improved performance compared to global sequence alignment and other existing methods on both prokaryotic and eukaryotic datasets.
  • The SVM approach showed resilience to sequence homology, unlike homology-based methods whose performance degrades with lower sequence identity.
  • A hybrid method combining the two-level SVM classifier and homology search was developed as a practical tool for sequence annotation.

Conclusions:

  • The developed two-level SVM system provides a highly accurate and reliable method for predicting protein subcellular localization.
  • This approach mitigates the bias introduced by homologous sequences in benchmark datasets, offering a more realistic performance assessment.
  • The hybrid method offers a versatile solution for annotating protein subcellular localization in bioinformatics.