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

A supervised hidden markov model framework for efficiently segmenting tiling array data in transcriptional and

Jiang Du1, Joel S Rozowsky, Jan O Korbel

  • 1Department of Computer Science, Yale University, New Haven, CT 06520, USA.

Bioinformatics (Oxford, England)
|October 14, 2006
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

HOROSCOPE: Decoding human centromere architecture from short reads using <i>k</i> -mer signatures.

bioRxiv : the preprint server for biology·2026
Same author

Population-scale Y chromosome assemblies reveal recurrent remodeling within constrained architectures.

bioRxiv : the preprint server for biology·2026
Same author

Recurrent DNA break clusters drive replication-stress-induced copy number variants and genome diversification.

Nature communications·2026
Same author

Population differences of chromosome 22q11.2 duplication structure predispose differentially to microdeletion and inversion.

Nature communications·2026
Same author

A snapshot of the earliest stages of hepatocellular carcinoma.

Cancer cell·2026
Same author

Efficient protein structure generation with sparse denoising models.

Nature machine intelligence·2025
This summary is machine-generated.

We developed a supervised framework using hidden Markov models (HMMs) for segmenting genomic tiling array data. This approach improves the identification of active genomic regions by incorporating biological knowledge, outperforming previous unsupervised methods.

Area of Science:

  • Genomics
  • Computational Biology
  • Bioinformatics

Background:

  • Large-scale tiling array experiments are crucial in genomics, particularly for projects like ENCODE.
  • Consistent segmentation of tiling array data into active regions (e.g., transfrags, ChIP-chip binding sites) is essential.
  • Previous unsupervised methods relied solely on signal distribution, lacking explicit biological knowledge integration.

Purpose of the Study:

  • To introduce a supervised framework for segmenting genomic tiling array data.
  • To leverage validated biological knowledge for improved accuracy in identifying active genomic regions.
  • To enable formal training and testing of segmentation models.

Main Methods:

  • Utilized a hidden Markov model (HMM) framework to model dependencies between neighboring probes.

Related Experiment Videos

  • Extended HMMs (generalized HMM) to describe state duration densities.
  • Defined a formal tiling-array analysis problem and developed strategies for sampling genomic regions for validation and creating gold-standard datasets for training and testing.

    • Main Results:

    • Demonstrated that the HMM framework efficiently processes tiling array data, performing comparably or better than existing methods.
    • Showcased how training data size and noise impact HMM performance on ENCODE transcriptional and ChIP-chip data.
    • Identified maximum entropy sampling as an optimal strategy for validation experiments, yielding the best-performing gold-standard datasets.

    Conclusions:

    • The supervised HMM framework offers a robust and accurate method for segmenting genomic tiling array data.
    • Incorporating validated biological knowledge significantly enhances the reliability of active region identification.
    • Optimal sampling strategies are critical for efficient and effective validation of large-scale genomic experiments.