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

Protein topology classification using two-stage support vector machines.

Jayavardhana Gubbi1, Alistair Shilton, Michael Parker

  • 1Department of Electrical and Electronics Engineering, The University of Melbourne, Parkville, Victoria 3010, Australia. jrgl@ee.unimelb.edu.au

Genome Informatics. International Conference on Genome Informatics
|May 16, 2007
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

How "America First" Abandoned Global Health: The Case for an African Model.

The American journal of bioethics : AJOB·2026
Same author

Minimum Foot Clearance Prediction in Stroke Survivors: A Transformer-Based Approach.

IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society·2026
Same author

Exploring the utility of artificial intelligence in identifying progression of prostate cancer during active surveillance: A systematic review.

Prostate cancer and prostatic diseases·2026
Same author

Four moral models of global health security.

PloS one·2026
Same author

Public engagement professionals: Exploring ethical tensions in communication, engagement and co-creation.

Public understanding of science (Bristol, England)·2026
Same author

Optimising rapid prenatal exome sequencing in the NHS genomic medicine service: the EXPRESS Synopsis.

Health and social care delivery research·2026
Same journal

Linear regression models predicting strength of transcriptional activity of promoters.

Genome informatics. International Conference on Genome Informatics·2012
Same journal

Sign: large-scale gene network estimation environment for high performance computing.

Genome informatics. International Conference on Genome Informatics·2012
Same journal

Docking-calculation-based method for predicting protein-RNA interactions.

Genome informatics. International Conference on Genome Informatics·2012
Same journal

Mechanism of cell cycle disruption by multiple p53 pulses.

Genome informatics. International Conference on Genome Informatics·2012
Same journal

Database for crude drugs and Kampo medicine.

Genome informatics. International Conference on Genome Informatics·2012
Same journal

A dynamic programming algorithm to predict synthesis processes of tree-structured compounds with graph grammar.

Genome informatics. International Conference on Genome Informatics·2011
See all related articles

Predicting protein 3-D structure from sequence is challenging for low sequence identity (<25%). This study introduces a two-stage support vector machine (SVM) approach to accurately classify protein structure classes and topologies, aiding molecular replacement in X-ray crystallography.

Area of Science:

  • Computational Biology
  • Structural Bioinformatics
  • Machine Learning in Biochemistry

Background:

  • Determining a protein's 3-D structure from its amino acid sequence is crucial for methods like molecular replacement in X-ray crystallography.
  • Homology modeling is effective for sequence identities >30%, but predicting structure for <25% sequence identity remains a significant challenge.

Purpose of the Study:

  • To develop a computational method for classifying protein structure topology and class for sequences with <25% identity.
  • To improve the accuracy of predicting protein structural classes and topologies, aiding structural determination.

Main Methods:

  • A two-stage support vector machine (SVM) approach was employed.
  • Input features included amino acid sequence, predicted secondary structure, and predicted relative solvent accessibility.

Related Experiment Videos

  • The first stage classified sequences into three structural classes (alpha, beta, alpha+beta); the second stage predicted one of 39 topologies.

    • Main Results:

    • The method achieved an overall accuracy of 87.44% for class prediction and 83.15% for topology prediction on a dataset with <25% pairwise sequence identity.
    • Class prediction yielded a sensitivity of 0.77 and a specificity of 0.91.
    • The SVM implementation (SVMHEAVY) and data are available for download.

    Conclusions:

    • The proposed two-stage SVM method effectively predicts protein structural classes and topologies even with low sequence identity.
    • This approach offers a valuable tool for tackling the phase problem in X-ray crystallography when traditional homology modeling is not feasible.
    • The study demonstrates the power of machine learning in addressing complex problems in structural bioinformatics.