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

Logistic regression for disease classification using microarray data: model selection in a large p and small n case.

J G Liao1, Khew-Voon Chin

  • 1Drexel University School of Public Health, Philadelphia, PA 19102, USA. jl544@drexel.edu

Bioinformatics (Oxford, England)
|June 2, 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

Gut microbiome and diet in populations with obesity: Role of the Na+/K+-ATPase transporter signaling in severe COVID-19.

Obesity (Silver Spring, Md.)·2022
Same author

Loss of Ceacam1 promotes prostate cancer progression in Pten haploinsufficient male mice.

Metabolism: clinical and experimental·2020
Same author

Prior robust empirical Bayes inference for large-scale data by conditioning on rank with application to microarray data.

Biostatistics (Oxford, England)·2013
Same author

Salinomycin, a polyether ionophoric antibiotic, inhibits adipogenesis.

Biochemical and biophysical research communications·2012
Same author

Survey of American food trends and the growing obesity epidemic.

Nutrition research and practice·2011
Same author

Prieurianin Causes Weight Loss in Diet-Induced Obese Mice and Inhibits Adipogenesis in Cultured Preadipocytes.

Journal of diabetes & metabolism·2011

A new parametric bootstrap model accurately estimates prediction error for microarray data. This method guides gene selection and shrinkage in penalized logistic regression, improving disease classification accuracy.

Area of Science:

  • Bioinformatics
  • Statistical Genetics
  • Computational Biology

Background:

  • Logistic regression is standard for binary outcomes but requires feature selection for high-dimensional microarray data.
  • Existing prediction error estimation methods are biased and variable when ignoring feature selection.
  • Penalized logistic regression with feature selection needs specialized statistical tools for model selection.

Purpose of the Study:

  • To develop a parametric bootstrap model for accurate prediction error estimation in microarray data analysis.
  • To provide guidance on optimal gene selection and shrinkage parameters for penalized logistic regression.
  • To improve disease classification models using high-dimensional genomic data.

Main Methods:

  • A parametric bootstrap approach is proposed, leveraging local false discovery rate principles.

Related Experiment Videos

  • The method addresses model selection challenges in two-step approaches (feature selection + penalized logistic regression).
  • Guidance is provided for determining the number of genes and optimal shrinkage in penalized logistic regression.

    • Main Results:

    • The proposed method offers more accurate prediction error estimation compared to generic methods.
    • Selecting over 20 genes typically yields minimal improvement in prediction error.
    • Highly accurate prediction models were achieved when applied to leukemia and cervical cancer datasets.

    Conclusions:

    • The parametric bootstrap model effectively addresses prediction error estimation in penalized logistic regression for microarray data.
    • The study offers practical insights into optimizing gene selection and shrinkage for robust disease classification.
    • Accurate prediction models for diseases like leukemia and cervical cancer can be developed using this approach.