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

At what scale should microarray data be analyzed?

Shuguang Huang1, Adeline A Yeo, Lawrence Gelbert

  • 1Genomic Informatics, Eli Lilly & Company, Indianapolis, Indiana 46285, USA. Huang_Shuguang@lilly.com

American Journal of Pharmacogenomics : Genomics-Related Research in Drug Development and Clinical Practice
|April 3, 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

Digital speech-based markers to advance prognosis in Alzheimer's disease.

Alzheimer's & dementia : the journal of the Alzheimer's Association·2026
Same author

Lateral Cervical Ectopic Thyroid: A Seven-Step Diagnostic Algorithm and Dual-Axis Treatment Strategy Derived From Case Analysis and Literature Review.

Head & neck·2026
Same author

Biomarkers in Alzheimer's disease clinical trials: 2025.

Alzheimer's & dementia (New York, N. Y.)·2026
Same author

The Right Person, the Right Treatment, at the Right Time in Alzheimer's Disease: Insights From the 2025 Brain Aging Symposium.

Aging cell·2025
Same author

The impact of recent approvals on future alzheimer's disease clinical development: Statistical considerations for combination trials.

The journal of prevention of Alzheimer's disease·2025
Same author

Key considerations for combination therapy in Alzheimer's clinical trials: Perspectives from an expert advisory board convened by the Alzheimer's drug discovery foundation.

The journal of prevention of Alzheimer's disease·2025
Same journal

Oncogenes as novel targets for cancer therapy (part IV): regulators of the cell cycle and apoptosis.

American journal of pharmacogenomics : genomics-related research in drug development and clinical practice·2005
Same journal

Using bioinformatics for drug target identification from the genome.

American journal of pharmacogenomics : genomics-related research in drug development and clinical practice·2005
Same journal

Antimycotic drug discovery in the age of genomics.

American journal of pharmacogenomics : genomics-related research in drug development and clinical practice·2005
Same journal

Cancer diagnostics: decision criteria for marker utilization in the clinic.

American journal of pharmacogenomics : genomics-related research in drug development and clinical practice·2005
Same journal

The Personalized Medicine Coalition: goals and strategies.

American journal of pharmacogenomics : genomics-related research in drug development and clinical practice·2005
Same journal

Expectations from structural genomics revisited: an analysis of structural genomics targets.

American journal of pharmacogenomics : genomics-related research in drug development and clinical practice·2005
See all related articles

Transformations like Box-Cox can stabilize microarray data variance when analyzing all genes together. For individual gene analysis, using the wrong scale slightly reduces statistical power but remains acceptable for moderate fold-changes.

Area of Science:

  • Bioinformatics
  • Statistical Genetics
  • Microarray Data Analysis

Background:

  • Microarray data hybridization intensities often require transformation to meet statistical model assumptions like normality and variance homogeneity.
  • Two common analysis strategies exist: pooled gene correlation analysis and gene-by-gene ANOVA.
  • Appropriate data scaling is crucial for reliable statistical modeling in microarray studies.

Purpose of the Study:

  • Investigate distributional properties of Affymetrix GeneChip signal data.
  • Evaluate the impact of data scaling on statistical analysis outcomes.
  • Compare different transformation methods for microarray data.

Main Methods:

  • Applied Box-Cox type transformations for pooled gene analysis.

Related Experiment Videos

  • Utilized log-transformation for comparative analysis.
  • Explored normality assumptions for gene-by-gene analysis.
  • Assessed the effects of log and quartic-root transformations on statistical power.

    • Main Results:

    • Box-Cox transformation effectively removed expression-variation dependency in pooled gene analysis.
    • Gene-by-gene analysis revealed gene-dependent intensity distributions.
    • Using an incorrect data scale resulted in some power loss, particularly with quartic-root transformation.
    • The t-test demonstrated robustness, making power loss acceptable for moderate fold-changes.

    Conclusions:

    • Data transformation is essential for appropriate statistical modeling of microarray data.
    • Box-Cox transformation is effective for stabilizing variance in pooled analyses.
    • While incorrect scaling impacts power in individual gene analysis, the effect is often manageable.
    • Choosing the right transformation scale is critical for maximizing statistical power in microarray studies.