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

Bayesian data integration: a functional perspective.

Curtis Huttenhower1, Olga G Troyanskaya

  • 1Department of Computer Science, Lewis-Sigler Institute for Integrative Genomics, Princeton University Princeton, NJ 08544, USA.

Computational Systems Bioinformatics. Computational Systems Bioinformatics Conference
|March 21, 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

Circulating imidazole propionate and coronary heart disease risk: interplay between histidine intake, fiber, and gut microbiome.

BMC medicine·2026
Same author

Item recognition is associated with gut microbiota composition in healthy humans.

Learning & memory (Cold Spring Harbor, N.Y.)·2026
Same author

Shotgun Metagenomic Profiling of the Gut Virome in Prodromal and Confirmed Parkinson's Disease.

Annals of neurology·2026
Same author

Long-lasting gut microbiome and fecal metabolome alterations after colorectal adenoma removal and their relationship to colorectal cancer.

Cell host & microbe·2026
Same author

Dietary sulfur amino acids enhance anti-tumor immunity in colon cancer via an NKT cell-XCL1-cDC1 circuit.

Immunity·2026
Same author

The relation between household dogs and cats during childhood/adolescence with social role outcomes and loneliness among youth with attention-deficit/hyperactivity disorder.

European child & adolescent psychiatry·2026
Same journal

Novel Gene Discovery in the Human Malaria Parasite using Nucleosome Positioning Data.

Computational systems bioinformatics. Computational Systems Bioinformatics Conference·2014
Same journal

Proceedings of Computational Systems Bioinformatics 2008. August 26-29, 2008. Palo Alto, California, USA.

Computational systems bioinformatics. Computational Systems Bioinformatics Conference·2009
Same journal

Graph wavelet alignment kernels for drug virtual screening.

Computational systems bioinformatics. Computational Systems Bioinformatics Conference·2009
Same journal

Fast multisegment alignments for temporal expression profiles.

Computational systems bioinformatics. Computational Systems Bioinformatics Conference·2009
Same journal

Knowledge representation and data mining for biological imaging.

Computational systems bioinformatics. Computational Systems Bioinformatics Conference·2009
Same journal

Efficient haplotype inference from pedigrees with missing data using linear systems with disjoint-set data structures.

Computational systems bioinformatics. Computational Systems Bioinformatics Conference·2009
See all related articles

Bayesian learning methods improve protein function prediction from genomic data. Performance varies across biological functions, highlighting the need for function-specific analysis in systems biology.

Area of Science:

  • Systems Biology
  • Bioinformatics
  • Computational Biology

Background:

  • Predicting protein function and interactions from genomic data is crucial for systems biology.
  • Integrating heterogeneous data sources presents challenges due to noise, coverage, and biases.
  • Understanding the robustness of data integration methods across different biological functions is essential.

Purpose of the Study:

  • To evaluate the effectiveness of Bayesian networks in predicting protein functional relationships.
  • To compare expert-estimated probabilities with machine learning approaches (generative and discriminative).
  • To analyze the impact of network structure and individual data sources on prediction accuracy across biological functions.

Main Methods:

  • Utilized Bayesian networks for predicting protein functional relationships.

Related Experiment Videos

  • Compared expert-defined conditional probabilities against models learned via expectation maximization (generative) and extended logistic regression (discriminative).
  • Assessed the contribution of individual data sources and interpreted results globally and for specific biological processes.

    • Main Results:

    • Learned models consistently outperformed expert-estimated models, especially with larger datasets.
    • Prediction accuracy varied significantly across different biological functions, with some categories being more challenging to predict.
    • Bayesian learning demonstrated a consistent benefit for data integration, but performance is context-dependent on functional categories.

    Conclusions:

    • Bayesian learning offers a robust approach for integrating heterogeneous genomic data for protein function prediction.
    • It is critical to consider functional specificity, as global performance metrics may mask variations in accuracy across biological categories.
    • The findings generalize to other data integration methods, emphasizing the importance of function-aware analysis in systems biology.