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

The ties problem resulting from counting-based error estimators and its impact on gene selection algorithms.

Xin Zhou1, K Z Mao

  • 1School of Electrical & Electronic Engineering, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore.

Bioinformatics (Oxford, England)
|August 16, 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

[Safety study of navigation-assisted medial "in-out-in" technique in C<sub>2</sub> screw fixation].

Zhonghua wai ke za zhi [Chinese journal of surgery]·2025
Same author

Improving robustness of gene ranking by multi-criterion combination with novel gene importance transformation.

International journal of data mining and bioinformatics·2013
Same author

Genome-wide association studies-derived susceptibility loci in type 2 diabetes: confirmation in a Chinese population.

Clinical and investigative medicine. Medecine clinique et experimentale·2012
Same author

Left ventricular systolic dyssynchrony in patients with isolated symptomatic myocardial bridge.

Scandinavian cardiovascular journal. Supplement·2012
Same author

Exogenous nerve growth factor supplementation elevates myocardial immunoreactivity and attenuates cardiac remodeling in pressure-overload rats.

Acta biochimica et biophysica Sinica·2012
Same author

A 90-day feeding study of glyphosate-tolerant maize with the G2-aroA gene in Sprague-Dawley rats.

Food and chemical toxicology : an international journal published for the British Industrial Biological Research Association·2012

Counting-based error estimators in gene selection for microarray data cause ties, leading to uncertainty. Continuous evaluation criteria offer improved gene selection by avoiding this issue.

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Genomics

Background:

  • Microarray data analysis commonly employs filter and wrapper methods for gene selection.
  • Wrapper methods typically use classification error as a feature subset evaluation criterion.
  • High dimensionality and small sample sizes in microarray data challenge the efficacy of counting-based error estimation for gene selection.

Purpose of the Study:

  • To investigate the limitations of counting-based error estimators in gene selection for microarray data.
  • To identify the causes of uncertainty in gene selection due to ties in error estimation.
  • To propose and evaluate continuous evaluation criteria for improved gene selection.

Main Methods:

  • Analysis of counting-based error estimators (resubstitution, leave-one-out, cross-validation, bootstrap) for gene selection.

Related Experiment Videos

  • Identification of the 'ties problem' caused by discrete error estimators.
  • Implementation and comparison of continuous evaluation criteria (generalized absolute value of w2, modified Relief's measure) with traditional methods.

    • Main Results:

    • Counting-based error estimators exhibit a severe 'ties problem', where multiple gene subsets receive identical scores, causing selection uncertainty.
    • The discrete nature of counting-based estimators is identified as the cause of the ties problem.
    • Continuous evaluation criteria demonstrate superior performance in gene selection compared to counting-based methods.

    Conclusions:

    • Continuous evaluation criteria are more effective than counting-based estimators for gene selection in high-dimensional, small-sample microarray data.
    • Avoiding the 'ties problem' through continuous metrics leads to more reliable gene selection.
    • The proposed continuous criteria offer a robust alternative for gene selection algorithms.