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

Effective dimensionality for principal component analysis of time series expression data.

Michael Hörnquist1, John Hertz, Mattias Wahde

  • 1Department of Science and Technology, Linköping University, SE-601 74, Norrköping, Sweden. micho@itn.liu.se

Bio Systems
|October 18, 2003
PubMed
Summary

Principal component analysis (PCA) can overemphasize noise in large gene expression datasets. A new method reveals that the true informative dimensions are far fewer than expected, especially in noisy biological time series.

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

Glassy dynamics near the interpolation transition in deep recurrent networks.

Physical review. E·2025
Same author

An interpretable method for automated classification of spoken transcripts and written text.

Evolutionary intelligence·2023
Same author

Sleepy drivers on a slippery road: A pilot study using a driving simulator.

Journal of sleep research·2021
Same author

Lidar-Camera Semi-Supervised Learning for Semantic Segmentation.

Sensors (Basel, Switzerland)·2021
Same author

Lowered Expression of Tumor Suppressor Candidate MYO1C Stimulates Cell Proliferation, Suppresses Cell Adhesion and Activates AKT.

PloS one·2016
Same author

Cumulants of Hawkes point processes.

Physical review. E, Statistical, nonlinear, and soft matter physics·2015

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Genomics

Background:

  • Large-scale gene expression data analysis is crucial for understanding biological systems.
  • Principal Component Analysis (PCA) is a common technique for dimensionality reduction in such data.
  • Uncritical application of PCA can lead to the analysis of noise rather than true biological signals.

Purpose of the Study:

  • To develop and apply a procedure for determining the effective dimensionality of gene expression data.
  • To assess the impact of noise and data characteristics on PCA results.
  • To provide a cautionary note on the interpretation of PCA in high-dimensional biological data.

Main Methods:

  • Utilized a newly developed procedure to estimate the effective dimensionality of a dataset.

Related Experiment Videos

  • Analyzed gene expression data from rat central nervous system development.
  • Evaluated the influence of noise levels and gene expression independence on dimensionality.

    • Main Results:

    • The effective dimensionality of the analyzed noisy time series data was significantly lower than anticipated.
    • Noise and the lack of independence among gene expression levels were identified as key factors reducing effective dimensionality.
    • Increasing measurements within a single time series did not prove to be a fruitful strategy for increasing dimensionality.

    Conclusions:

    • The effective dimensionality of noisy biological time series data is often substantially lower than the number of measurements suggests.
    • Careful assessment of data noise and independence is essential before and during PCA application.
    • The findings underscore the need for robust methods to interpret high-dimensional biological data and avoid spurious conclusions from PCA.