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

Hybrid clustering for microarray image analysis combining intensity and shape features.

Jörg Rahnenführer1, Daniel Bozinov

  • 1Max-Planck-Institute for Informatics, Stuhlsatzenhausweg 85, D-66123 Saarbrücken, Germany. rahnenfj@mpi-sb.mpg.de

BMC Bioinformatics
|May 1, 2004
PubMed
Summary

This study introduces a novel hybrid algorithm for microarray image analysis, combining histogram and shape features for improved spot detection and quantification. The method enhances accuracy by effectively filtering artifacts and segmenting pixels into foreground, background, and deletions.

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

Ensuring Quality in Preclinical Research: The Importance of Being Human.

Biometrical journal. Biometrische Zeitschrift·2026
Same author

A review of machine learning in toxicology: current practices and reporting gaps.

Archives of toxicology·2026
Same author

Stage-dependent effects of systemic ASBT inhibition in a cholestasis-induced cholemic nephropathy mouse model.

JHEP reports : innovation in hepatology·2025
Same author

Elimination of CDX2 restricts intestinal hybrid differentiation signatures in stem cell-derived hepatocyte-like cells.

Stem cell research & therapy·2025
Same author

Adjustment for day-to-day variability in the estimation of effective concentrations for the assessment of mixture toxicity.

Archives of toxicology·2025
Same author

Critical discussion of recently published claims of a causal relationship between bat decline and infant mortality.

Archives of toxicology·2025

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Genomics

Background:

  • Microarray image analysis is critical for reliable experimental results.
  • Current methods rely on either spot shape or intensity histogram approaches.
  • A need exists for hybrid algorithms combining the strengths of both methods.

Purpose of the Study:

  • To develop a robust and adaptive hybrid method for microarray image analysis.
  • To integrate spot shape information into a pixel clustering approach.
  • To improve the accuracy of spot detection and quantification.

Main Methods:

  • Utilized pixel clustering, a histogram-based technique.
  • Integrated spot shape information by constructing a bivalence mask based on clustering results.

Related Experiment Videos

  • Applied the mask to filter data and refine cluster algorithm performance.
  • Defined and evaluated a 'stability' quality measure.

    • Main Results:

    • Demonstrated effective integration of spot shape into pixel clustering for microarray analysis.
    • The hybrid approach improved cluster algorithm results through shape-based filtering.
    • A real data set was used to evaluate the 'stability' measure.
    • Comparison with established 'Spot' software on replicate data showed improved performance.

    Conclusions:

    • The developed method offers a successful hybrid solution for microarray image analysis.
    • It effectively combines histogram and shape features, adapted for microarray image characteristics.
    • The filtering step categorizes pixels into foreground, background, and deletions for artifact elimination.