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

Detecting differential gene expression with a semiparametric hierarchical mixture method.

Michael A Newton1, Amine Noueiry, Deepayan Sarkar

  • 1Department of Statistics, University of Wisconsin-Madison, 1210 West Dayton St, Madison, WI 53706-1685, USA. newton@stat.wisc.edu

Biostatistics (Oxford, England)
|April 1, 2004
PubMed
Summary

This study introduces a flexible hierarchical mixture model for analyzing microarray data, improving the detection of differential gene expression. The new method offers enhanced sensitivity and accuracy in identifying gene expression changes.

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

Microbiological spectrum and probability of transmission of infection from ocular tissues of enucleated eyes harvested from septicemic donors.

World journal of transplantation·2026
Same author

Genetic complementation reveals structure-function links in nodavirus RNA replication complex crowns.

PLoS pathogens·2026
Same author

Nodavirus protein A's interdomain elbow controls RNA replication organelle formation and function.

Nucleic acids research·2026
Same author

Thioredoxin interacting protein (TXNIP), a redox regulator, mediates the RAPGEF3/4 signaling dependency in primary melanoma.

bioRxiv : the preprint server for biology·2026
Same author

SatTCR: a pipeline for performing saturation analysis of the T cell receptor repertoire and a case study of a healthy canine.

MethodsX·2025
Same author

Discovery and Targeted Proteomic Studies Reveal Striatal Markers Validated for Huntington's Disease.

Annals of clinical and translational neurology·2025

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Statistical Genetics

Background:

  • Mixture modeling is effective for analyzing differential gene expression in microarray data.
  • Fully parametric models may lack flexibility, while existing flexible models may lack sensitivity with limited data.
  • There is a need for more sensitive and flexible models to capture complex variability in normalized microarray data.

Purpose of the Study:

  • To propose a hierarchical mixture model for sensitive and flexible detection of differential gene expression.
  • To address limitations of existing parametric and semiparametric mixture models in microarray analysis.
  • To provide a robust statistical framework for the two-sample comparison problem in gene expression studies.

Main Methods:

  • Development of a hierarchical mixture model incorporating both parametric and semiparametric approaches.

Related Experiment Videos

  • Utilizing expectation-maximization (EM)-based algorithms for model fitting.
  • Application to the two-sample comparison problem using Affymetrix microarray data from yeast translation experiments.

    • Main Results:

    • The proposed hierarchical mixture model demonstrates good operating characteristics in simulations and real data analyses.
    • The methodology provides sensitive detection of differential expression, even with limited measurements per gene.
    • Gene-specific posterior probabilities effectively define gene lists and control false discovery rates.

    Conclusions:

    • The hierarchical mixture model offers a sensitive and flexible approach for differential gene expression analysis in microarray data.
    • This methodology effectively accounts for complex variability in normalized microarray data.
    • The proposed method outperforms competing methodologies in simulation, spike-in, and cross-validation analyses.