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

Performance of variable selection methods using stability-based selection.

Danny Lu1, Aalim Weljie2, Alexander R de Leon3

  • 1Sick Kids Research Institute, 555 University Avenue, Toronto, ON, M5G 1X8, Canada.

BMC Research Notes
|April 6, 2017
PubMed
Summary

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

Vaccination elicits HIV broadly neutralizing antibodies in primates.

Nature·2026
Same author

Rapid boosting increases germinal center responses to sequential vaccines.

Nature immunology·2026
Same author

A tumor-intrinsic WNT-inhibitory NOTUM program drives immune resistance in microsatellite stable colorectal cancer.

Cell reports. Medicine·2026
Same author

Structural and immunogenetic signatures guide CD4-mimetic HIV vaccine development.

Cell reports·2026
Same author

Deep learning-enabled scaffolding of spatial arrays of PfCSP epitopes.

Proceedings of the National Academy of Sciences of the United States of America·2026
Same author

Local antibody feedback enforces a checkpoint on affinity maturation in the germinal center and promotes epitope spreading.

Immunity·2026
This summary is machine-generated.

Variable selection in high-dimensional data, like metabolomics, benefits from stability selection. Variable Importance in Projection (VIP) scores and FDR-adjusted t-tests offer robust predictive performance, especially with BioMark R package.

Area of Science:

  • High-dimensional data analysis
  • Metabolomics
  • Bioinformatics

Background:

  • Variable selection is crucial for high-dimensional data analysis, particularly in metabolomics.
  • This study evaluates four variable selection methods, incorporating stability-based selection and false discovery rate (FDR) adjustment.

Purpose of the Study:

  • To compare the predictive performance of different variable selection methods in high-dimensional datasets.
  • To assess the utility of stability-based selection and FDR adjustment in metabolomics studies.

Main Methods:

  • Utilized simulation studies with varying biological data factors.
  • Employed stability-based selection implemented in the R package BioMark.
  • Evaluated methods using partial area under the receiver operating characteristic curve (pAUC).
Keywords:
False discovery rate (FDR)High-dimensional biological dataPartial area under the receiver-operating characteristic curve (pAUC)Stability-based variable selectionVariable importance in projection (VIP)

Related Experiment Videos

Main Results:

  • No single method excelled in all scenarios; however, stability-selected t-tests (with or without FDR) and VIP scores from stability-selected PLS models showed strong performance.
  • Real metabolomics data yielded similar findings.
  • Group sample size, effect size, number of significant variables, and correlation structure were key factors influencing performance.

Conclusions:

  • Variable Importance in Projection (VIP) scores with stability selection are recommended for small to modest variable sets.
  • The FDR-adjusted t-test performed best for high variable counts with block correlation.
  • The BioMark R package offers an effective and user-friendly tool for variable selection.