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

Gene selection for sample classifications in microarray experiments.

Chen-An Tsai1, Chun-Houh Chen, Te-Chang Lee

  • 1Division of Biometry and Risk Assessment, National Center for Toxicological Research, Food and Drug Administration, Jefferson, Arkansas 72079, USA.

DNA and Cell Biology
|December 9, 2004
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

The Taiwan Precision Medicine Initiative provides a cohort for large-scale studies.

Nature·2025
Same author

Population-specific polygenic risk scores for people of Han Chinese ancestry.

Nature·2025
Same author

Response to comment on "Alteration of gut microbial composition associated with the therapeutic efficacy of fecal microbiota transplantation in Clostridium difficile infection".

Journal of the Formosan Medical Association = Taiwan yi zhi·2025
Same author

Clinical impact of pharmacogenetic risk variants in a large chinese cohort.

Nature communications·2025
Same author

Identifying Distinct Developmental Patterns of Brain Complexity in Autism: A Cross-Sectional Cohort Analysis Using the Autism Brain Imaging Data Exchange.

Psychiatry and clinical neurosciences·2025
Same author

PIAS1 S510G variant acts as a genetic modifier of spinocerebellar ataxia type 3 by selectively impairing mutant ataxin-3 proteostasis.

The international journal of biochemistry & cell biology·2024
Same journal

Fibroblast Growth Factor 18 Regulates Mitochondrial Function to Alleviate Myocardial Ischemia-Reperfusion Injury in Rats.

DNA and cell biology·2026
Same journal

Silencing Long Noncoding RNA <i>CRNDE</i> Alleviates Hypoxia-Induced Myocardial Cell Injury by Inhibiting the NLRP3/ASC Pathway.

DNA and cell biology·2026
Same journal

Integrated Multi-Omics Profiling Reveals Coordinated Transcriptional, Translational, and Metabolic Remodeling in IPF Fibroblasts.

DNA and cell biology·2026
Same journal

The Innate Immune Protein IFITM3 as a γ-Secretase Modulatory Protein: From Inflammation to Alzheimer's Disease.

DNA and cell biology·2026
Same journal

Glycosyltransferases as Oncogenic Drivers: Lessons from Cancer Genome Mining.

DNA and cell biology·2026
Same journal

Essential Role for XRCC1 and PARP1 in the Repair of Genotoxic Damage During Spermatogenesis.

DNA and cell biology·2026
See all related articles

This study introduces a family-wise error rate approach to select important genes for classifying samples using DNA microarrays. This method improves classification accuracy for cancer and toxicology studies by reducing noise from irrelevant genes.

Area of Science:

  • Genomics
  • Bioinformatics
  • Statistical Genetics

Background:

  • DNA microarrays enable global gene expression profiling but face challenges with high dimensionality (many genes, few samples).
  • Using all genes can lead to reduced classification performance due to noise from non-discriminatory genes.
  • Optimal gene subset selection is crucial for accurate sample classification.

Purpose of the Study:

  • To propose and evaluate a novel family-wise error (FWE) rate approach for selecting discriminatory genes in sample classification.
  • To assess the performance of the FWE approach in two-sample and multiple-sample classification scenarios.
  • To compare the proposed method with existing techniques using public colon cancer and toxicogenomic datasets.

Main Methods:

  • A family-wise error (FWE) rate approach was developed to control false positives during gene selection.

Related Experiment Videos

  • The proposed method was applied to a colon cancer dataset using k-nearest neighbors (k-NN) and support vector machine (SVM) classifiers.
  • The FWE approach was also tested on a toxicogenomic dataset with nine treatments (control and eight metals) for multi-sample classification.

    • Main Results:

    • The FWE-selected gene sets demonstrated improved or comparable classification performance to existing methods on the colon cancer dataset.
    • For the toxicogenomic dataset, SVM classification using FWE-selected genes achieved >80% accuracy for internal cross-validation and >70% for external cross-validation.
    • The ANOVA F-test gene set (omegaF) showed slightly higher accuracy than the t-test gene set (omegaT) in internal cross-validation, with both performing equally in external cross-validation.

    Conclusions:

    • The proposed family-wise error rate approach is effective for selecting discriminatory genes in high-dimensional microarray data.
    • This method enhances classification accuracy in both two-sample and multi-sample settings, applicable to cancer and toxicogenomics research.
    • The FWE approach offers a robust strategy for gene subset selection, outperforming or matching existing univariate methods.