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

Domain prediction with probabilistic directional context.

Alejandro Ochoa1,2, Mona Singh1,3

  • 1Lewis-Sigler Institute for Integrative Genomics.

Bioinformatics (Oxford, England)
|April 14, 2017
PubMed
Summary

We developed a new probabilistic method for protein domain prediction that considers the order of domains. This approach improves prediction accuracy by approximately 15% compared to existing methods.

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

FBApro: A fast, simple linear transformation for diverse metabolic modeling tasks.

ArXiv·2026
Same author

Fast, accurate construction of multiple sequence alignments from protein language embeddings.

bioRxiv : the preprint server for biology·2026
Same author

Single-cell transcriptomics reveals FXR1 as an actionable target for siRNA therapy in ovarian cancer.

Nature communications·2026
Same author

Polygenic risk scores and HLA class II variants are biomarkers of corticosteroid response in childhood nephrotic syndrome.

Kidney international·2026
Same author

Preface.

Journal of computational biology : a journal of computational molecular cell biology·2026
Same author

Kinship estimation bias carries over to heritability estimation bias using variance components.

Genetics·2025

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Genomics

Background:

  • Protein domain prediction is crucial for understanding protein function from sequence data.
  • Current methods often predict domains independently, overlooking co-occurrence patterns and order.
  • Existing approaches that consider domain co-occurrence do not model order probabilistically.

Purpose of the Study:

  • To introduce a novel probabilistic approach for protein domain prediction that incorporates directional domain context.
  • To develop a method that scores all domain pairs within a sequence, accounting for their order.
  • To improve the accuracy and completeness of protein domain predictions.

Main Methods:

  • Developed a probabilistic model for 'directional' domain context.

Related Experiment Videos

  • Extended a previous Markov model-based approach to include pairwise terms.
  • Formulated the optimization problem as an integer linear program solvable in practice.
  • Implemented the approach as dPUC2 (Domain Prediction Using Context).

    • Main Results:

    • The method scores all domain pairs, considering their order, even for non-sequential domains.
    • The approach can be interpreted within the framework of Markov random fields.
    • Incorporating domain context increased predictions by approximately 15%.
    • The dPUC2 approach outperformed all competing methods in extensive evaluations.

    Conclusions:

    • Probabilistic modeling of directional domain context significantly enhances protein domain prediction.
    • The dPUC2 method offers a robust and efficient solution for sequence-based function prediction.
    • This work provides a new standard for incorporating domain co-occurrence information in prediction tasks.