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

SVM-BALSA: remote homology detection based on Bayesian sequence alignment.

Bobbie-Jo Webb-Robertson1, Christopher Oehmen, Melissa Matzke

  • 1Computational Biology and Bioinformatics, Pacific Northwest National Laboratory, Richland, WA 99352, USA. bj@pnl.gov

Computational Biology and Chemistry
|November 18, 2005
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

Low Peripheral Blood Counts and Elevated Proinflammatory Cytokines Signal a Poor CD19 Chimeric Antigen Receptor T-cell Response in Acute Lymphoblastic Leukemia.

Transplantation and cellular therapy·2025
Same author

Protection of β cells against pro-inflammatory cytokine stress by the GDF15-ERBB2 signaling.

medRxiv : the preprint server for health sciences·2023
Same author

Inhibition of polyamine biosynthesis preserves β cell function in type 1 diabetes.

Cell reports. Medicine·2023
Same author

Identification of a specific exporter that enables high production of aconitic acid in Aspergillus pseudoterreus.

Metabolic engineering·2023
Same author

Systematic review of type 1 diabetes biomarkers reveals regulation in circulating proteins related to complement, lipid metabolism, and immune response.

Clinical proteomics·2023
Same author

Systematic review of type 1 diabetes biomarkers reveals regulation in circulating proteins related to complement, lipid metabolism, and immune response.

medRxiv : the preprint server for health sciences·2023

This study introduces SVM-BALSA, a novel bioinformatics tool using Bayesian alignment scores to enhance protein homology detection. SVM-BALSA improves sensitivity in identifying evolutionarily related proteins compared to traditional methods.

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Protein Science

Background:

  • Biopolymer sequence comparison is crucial for identifying homologous proteins.
  • Support Vector Machines (SVMs) offer a statistical learning approach for protein family classification.
  • Existing SVM methods show promise, outperforming iterative profile methods like PSI-BLAST.

Purpose of the Study:

  • To enhance protein homology detection using a Bayesian alignment score within SVMs.
  • To improve the sensitivity of identifying evolutionarily related proteins.
  • To evaluate the performance of the new SVM-BALSA algorithm against traditional methods.

Main Methods:

  • Developed the SVM-BALSA algorithm incorporating a Bayesian alignment score.
  • Accounted for the uncertainty of all possible alignments in SVM construction.

Related Experiment Videos

  • Utilized a benchmark dataset of 54 unique protein families for evaluation.

    • Main Results:

    • The SVM-BALSA algorithm demonstrated improved sensitivity over traditional dynamic programming implementations.
    • Achieved a higher area under the receiver operating characteristic (ROC) curves for 37 out of 54 protein families.
    • Showcased an improved overall performance curve at a significance level of 0.07.

    Conclusions:

    • Incorporating Bayesian alignment scores into SVMs significantly enhances protein homology detection.
    • SVM-BALSA offers a more sensitive and effective approach for identifying homologous proteins.
    • The findings suggest a new standard for computational methods in evolutionary protein analysis.