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

Estimation of false discovery proportion under general dependence.

Yudi Pawitan1, Stefano Calza, Alexander Ploner

  • 1Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden. yudi.pawitan@ki.se

Bioinformatics (Oxford, England)
|October 19, 2006
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

Effects of family genetic risk scores and environmental factors on risk of schizophrenia and bipolar disorder.

Molecular psychiatry·2026
Same author

Human Papillomavirus-Related Cancer in People With HIV and Solid Organ Transplant Recipients.

JAMA network open·2026
Same author

Preoperative Manual Detorsion in Children With Intravaginal Testicular Torsion: A Single-Center Experience.

World journal of surgery·2026
Same author

Polygenic overlap and shared genomic loci between anorexia nervosa and cardiometabolic traits suggest shared biological mechanisms.

Neurobiology of disease·2026
Same author

Genetic variation in antidiabetic drug targets: associations with Parkinson's disease risk and age at onset.

NPJ Parkinson's disease·2026
Same author

Humoral and Cellular Immune Response in Patients with Hematological Disorders After Three Doses of mRNA COVID-19 Vaccine: A Single-Center Observational Study.

Vaccines·2026

Gene expression data correlations can inflate false discovery rate (FDR) estimates. This study introduces an improved method, Empirical Likelihood of False Discovery (ELF), to accurately estimate the false discovery proportion (FDP) in correlated data.

Area of Science:

  • Bioinformatics
  • Statistical Genetics
  • Computational Biology

Background:

  • Gene expression data frequently exhibits correlations due to biological and technical factors.
  • These correlations can significantly increase the variability of standard false discovery rate (FDR) estimates.
  • The false discovery proportion (FDP) is a more appropriate metric for assessing differential expression in microarray data.

Purpose of the Study:

  • To address the impact of gene correlations on FDR estimation in microarray data.
  • To introduce and validate a novel method for estimating the false discovery proportion (FDP) that accounts for correlations.
  • To improve the accuracy of differential expression analysis in the presence of correlated gene expression data.

Main Methods:

  • Analysis of test statistic distribution variations using singular value decomposition (SVD).

Related Experiment Videos

  • Development of a latent FDR model that incorporates correlation effects.
  • Implementation of the Empirical Likelihood of False Discovery (ELF) estimation procedure using Poisson regression.

    • Main Results:

    • The proposed ELF method demonstrates substantially improved FDP estimation compared to standard FDR approaches for simulated correlated data.
    • ELF accurately estimates FDP in datasets with correlation structures derived from real-world gene expression data.
    • The study illustrates the application of ELF in analyzing breast cancer and lymphoma gene expression datasets.

    Conclusions:

    • The ELF method provides a more statistically robust approach to estimating false discovery rates in the presence of gene correlations.
    • Accurate FDP estimation using ELF enhances the reliability of differential gene expression findings from microarray studies.
    • The ELF procedure offers a valuable tool for researchers analyzing complex gene expression datasets.