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

Threshold gradient descent method for censored data regression with applications in pharmacogenomics.

J Gui1, H Li

  • 1Department of Statistics and Rowe Program in Genetics, University of California, Davis, CA 95616, USA. jgui@ucdavis.edu

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing
|March 12, 2005
PubMed
Summary

This study introduces a novel Threshold Gradient Descent (TGD) method for pharmacogenomics, effectively analyzing patient survival data. The TGD method aids in identifying crucial genes linked to cancer survival and building predictive models.

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

Phenobarbital induction mediated by a distal CYP2B2 sequence in rat liver transiently transfected in situ.

The Journal of biological chemistry·1996
Same author

Induction of phosphoglycerate kinase 1 gene expression by hypoxia. Roles of Arnt and HIF1alpha.

The Journal of biological chemistry·1996
Same author

Identification of novel regions of deletion in familial Wilms' tumor by comparative genomic hybridization.

Cancer research·1996
Same author

Alternatively spliced cyclin C mRNA is widely expressed, cell cycle regulated, and encodes a truncated cyclin box.

Oncogene·1996
Same author

Emergence of preferred structures in a simple model of protein folding.

Science (New York, N.Y.)·1996
Same author

Developmental regulation of collagen differential expression in the rabbit bladder.

The Journal of urology·1996

Area of Science:

  • Genomics
  • Pharmacogenomics
  • Biostatistics

Background:

  • Relating high-dimensional genetic data to clinical phenotypes is crucial in pharmacogenomics.
  • Survival phenotypes, accounting for censored data, offer more information than categorical variables.
  • High-dimensional and low-sample size data present computational challenges in genomic analysis.

Purpose of the Study:

  • To develop a Threshold Gradient Descent (TGD) method for Cox regression.
  • To select genes relevant to patient survival.
  • To build a predictive model for patient risk assessment.

Main Methods:

  • Developed a Threshold Gradient Descent (TGD) method tailored for Cox models.
  • Utilized gradient descent iterations to address computational challenges in high-dimensional, low-sample size settings.

Related Experiment Videos

  • Applied the TGD method to a real-world dataset for predicting survival in diffuse large B-cell lymphoma patients.

    • Main Results:

    • The TGD method successfully identified genes associated with cancer-related survival.
    • A parsimonious predictive model for patient survival was constructed.
    • TGD-based Cox regression demonstrated superior predictive performance compared to L2 penalized regression.
    • TGD outperformed L1 penalized regression in selecting relevant genes.

    Conclusions:

    • The TGD method is effective for gene selection and survival prediction in pharmacogenomics.
    • The approach addresses computational difficulties in high-dimensional genomic data analysis.
    • This method enhances the identification of key genes influencing patient survival and improves predictive modeling accuracy.