Incorporating higher-order representative features improves prediction in network-based cancer prognosis analysis

Public Deposited
Analytics

Downloadable Content

Download PDF
Request Version for Screen Reader
Creator
  • Ma, Shuangge
    • Other Affiliation: School of Public Health, Yale University, New Haven, CT, USA
  • Kosorok, Michael
    • Affiliation: Gillings School of Global Public Health, Department of Biostatistics
  • Huang, Jian
    • Other Affiliation: Departments of Statistics and Actuarial Science, and Biostatistics, University of Iowa, Iowa City, IA, USA
  • Dai, Ying
    • Other Affiliation: School of Public Health, Yale University, New Haven, CT, USA
Abstract
  • Abstract Background In cancer prognosis studies with gene expression measurements, an important goal is to construct gene signatures with predictive power. In this study, we describe the coordination among genes using the weighted coexpression network, where nodes represent genes and nodes are connected if the corresponding genes have similar expression patterns across samples. There are subsets of nodes, called modules, that are tightly connected to each other. In several published studies, it has been suggested that the first principal components of individual modules, also referred to as "eigengenes", may sufficiently represent the corresponding modules. Results In this article, we refer to principal components and their functions as representative features". We investigate higher-order representative features, which include the principal components other than the first ones and second order terms (quadratics and interactions). Two gradient thresholding methods are adopted for regularized estimation and feature selection. Analysis of six prognosis studies on lymphoma and breast cancer shows that incorporating higher-order representative features improves prediction performance over using eigengenes only. Simulation study further shows that prediction performance can be less satisfactory if the representative feature set is not properly chosen. Conclusions This study introduces multiple ways of defining the representative features and effective thresholding regularized estimation approaches. It provides convincing evidence that the higher-order representative features may have important implications for the prediction of cancer prognosis.
Date of publication
DOI
Identifier
Resource type
  • Article
Rights statement
  • In Copyright
Rights holder
  • Shuangge Ma et al.; licensee BioMed Central Ltd.
License
Journal title
  • BMC Medical Genomics
Journal volume
  • 4
Journal issue
  • 1
Page start
  • 5
Language
  • English
Is the article or chapter peer-reviewed?
  • Yes
ISSN
  • 1755-8794
Bibliographic citation
  • BMC Medical Genomics. 2011 Jan 12;4(1):5
Publisher
  • BioMed Central Ltd
Access right
  • Open Access
Date uploaded
  • August 23, 2012

Relations

Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
file details 1755-8794-4-5.pdf 2019-05-07 Public Download
file details Analysis results. This additional file contains the detailed analysis results for dataset D1-D6. 2019-05-07 Public Download
file details Results on network module construction. This additional file contains the details on the network modules constructed using WGCNA. 2019-05-07 Public Download
file details 1755-8794-4-5.xml 2019-05-07 Public Download