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 comparison of dimensionality reduction methods on RNA-Seq data from the GTEx project.

Ho-Sik Seok1

  • 1Department of Computer and Communications Engineering, Kangwon National University, Chuncheon-si, Gangwon-do, 24341, South Korea. hsseok@kangwon.ac.kr.

Genes & Genomics
|December 14, 2019
PubMed
Summary

Dimensionality reduction methods like LLE, t-SNE, and SE offer superior clustering performance on RNA-Seq data compared to PCA and MDS. Applying both linear and non-linear techniques is recommended for intuitive data understanding.

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

Enhancing performance of gene expression value prediction with cluster-based regression.

Genes & genomics·2021
See all related articles

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Genomics

Background:

  • Bioinformatics datasets often feature a high dimensionality (many features) with limited samples, hindering intuitive comprehension.
  • Dimensionality reduction and manifold learning offer solutions, but method selection remains crucial.

Purpose of the Study:

  • To evaluate the performance of various dimensionality reduction techniques on RNA-Seq data.
  • To compare linear methods (PCA, MDS) with non-linear methods (LLE, SE, t-SNE).

Main Methods:

  • Applied dimensionality reduction methods including Locally Linear Embedding (LLE), Multi-dimensional Scaling (MDS), Principal Component Analysis (PCA), Spectral Embedding (SE), and t-distributed Stochastic Neighbor Embedding (t-SNE).
  • Utilized RNA-Seq data from the Genotype-Tissue Expression (GTEx) project.
Keywords:
ClusteringDimensionality reductionManifold learningRNA-SeqThe genotype-tissue expression (GTEx) project

Related Experiment Videos

  • Assessed performance using k-means clustering in reduced dimensions (2, 3, and 4).

    • Main Results:

    • Each dimensionality reduction method generated distinct reduced spaces.
    • Non-linear methods (LLE, t-SNE, SE) outperformed linear methods (PCA, MDS) in k-means clustering accuracy.
    • Visualization revealed unique characteristics for each method in the reduced space.

    Conclusions:

    • Non-linear dimensionality reduction techniques show promise for analyzing complex bioinformatics data.
    • Combining linear and non-linear methods can enhance intuitive understanding of high-dimensional datasets.