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

Partial least squares dimension reduction for microarray gene expression data with a censored response.

Danh V Nguyen1

  • 1Division of Biostatistics, Public Health Sciences, School of Medicine, University of California, One Shields Avenue, Davis, CA 95616-8638, USA. ucdnguyen@ucdavis.edu

Mathematical Biosciences
|February 1, 2005
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

Nondialytic Care vs Dialysis Transition on Hospitalization: Outcomes in Veterans With Advanced Chronic Kidney Disease.

Mayo Clinic proceedings·2026
Same author

Continuous Glucose Monitoring and Mortality Risk Among U.S. Veterans Receiving Dialysis With Diabetes.

Diabetes care·2026
Same author

A pilot study of neuropsychological testing in a civilian traumatic brain injury clinic: Neurocognitive profiles, temporal trends and cognitive screening.

Applied neuropsychology. Adult·2026
Same author

Effects of Extracorporeal Life Support Volume on Neonatal Mortality in Congenital Diaphragmatic Hernia: A Retrospective Cohort Study.

The Journal of pediatrics·2026
Same author

Continuous glucose monitoring in patients with chronic kidney disease on dialysis and with kidney transplantation.

Current opinion in nephrology and hypertension·2026
Same author

Dietary potassium and fiber intake, survival, and health-related quality of life, in a prospective multi-center hemodialysis cohort.

Clinical nutrition (Edinburgh, Scotland)·2026
Same journal

The stability and bifurcations of ecosystems within resource constraints - Dedicated to Professor Shigui Ruan on the occasion of his 60th birthday.

Mathematical biosciences·2026
Same journal

The hydra and hormetic effects in a single discrete-time overcompensation model.

Mathematical biosciences·2026
Same journal

Seasonal impacts on brucellosis transmission mediated by live sheep supply-demand dynamics.

Mathematical biosciences·2026
Same journal

Optimal controls and cost-effectiveness analysis on the transmission dynamics of early blight disease in tomatoes.

Mathematical biosciences·2026
Same journal

Temperature-dependent dynamics and allee effect thresholds mediate fourfold cusp stability in biological control of invasive vectors.

Mathematical biosciences·2026
Same journal

Dynamics of a stochastic tumor-immune interaction system with an Ornstein-Uhlenbeck process.

Mathematical biosciences·2026
See all related articles

Partial Least Squares (PLS) outperformed Principal Component Analysis (PCA) for reducing dimensionality in cancer gene expression data, especially with low to moderate variance explained. PLS-PH showed the best overall performance despite high censoring rates.

Area of Science:

  • Genomics
  • Bioinformatics
  • Cancer Research

Background:

  • DNA microarrays monitor transcriptional programs in tumor cells for cancer research.
  • Comparing clinical outcomes (e.g., survival) with gene expression patterns is crucial.
  • High-dimensionality and censored survival times pose challenges in analyzing microarray data.

Purpose of the Study:

  • To systematically assess a two-stage procedure for relating gene expression to survival.
  • To compare Partial Least Squares (PLS) with Principal Component Analysis (PCA) and modified PLS (MPLS) for dimensionality reduction.
  • To examine the impact of censoring on the performance of these methods.

Main Methods:

  • A two-stage procedure combining PLS for dimensionality reduction and proportional hazard (PH) regression for survival prediction.

Related Experiment Videos

  • Simulation model using gene expression data with a censored response variable.
  • Comparison of PLS, PCA, and MPLS for the dimensionality reduction stage.

    • Main Results:

    • PLS demonstrated superior performance over PCA when predictor variance explained was low to moderate (40%-60%).
    • PLS performance was comparable to MPLS, with slight advantages in some scenarios.
    • All methods showed performance degradation with high censoring rates, but PLS-PH remained the best overall.

    Conclusions:

    • The PLS-based two-stage procedure is effective for analyzing high-dimensional gene expression data with censored survival times.
    • PLS is a more suitable dimensionality reduction technique than PCA for this type of data.
    • The PLS-PH approach offers robust performance even under significant data censoring.