Spike and slab gene selection for multigroup microarray data

Research output: Contribution to journalArticle

54 Citations (Scopus)

Abstract

DNA microarrays can provide insight into genetic changes that characterize different stages of a disease process. Accurate identification of these changes has significant therapeutic and diagnostic implications. Statistical analysis for multistage (multigroup) data is challenging, however. ANOVA-based extensions of two-sample Z-tests, a popular method for detecting differentially expressed genes in two groups, do not work well in multigroup settings. False detection rates are high because of variability of the ordinary least squares estimators and because of regression to the mean induced by correlated parameter estimates. We develop a Bayesian rescaled spike and slab hierarchical model specifically designed for the multigroup gene detection problem. Data preprocessing steps are introduced to deal with unique features of microarray data and to enhance selection performance. We show theoretically that spike and slab models naturally encourage sparse solutions through a process called selective shrinkage. This translates into oracle-like gene selection risk performance compared with ordinary least squares estimates. The methodology is illustrated on a large microarray repository of samples from different clinical stages of metastatic colon cancer. Through a functional analysis of selected genes, we show that spike and slab models identify important biological signals while minimizing biologically implausible false detections.

Original languageEnglish
Pages (from-to)764-780
Number of pages17
JournalJournal of the American Statistical Association
Volume100
Issue number471
DOIs
StatePublished - Sep 1 2005
Externally publishedYes

Fingerprint

Gene Selection
Microarray Data
Spike
Gene
Ordinary Least Squares Estimator
Data Preprocessing
DNA Microarray
Least Squares Estimate
Ordinary Least Squares
Functional Analysis
Hierarchical Model
Shrinkage
Microarray
Repository
Statistical Analysis
Diagnostics
Cancer
Regression
Methodology
Model

Keywords

  • Colon cancer
  • Hypervariance
  • Penalization
  • Rescaling
  • Risk misclassification
  • Shrinkage
  • Sparsity
  • Stochastic variable selection
  • Zcut

ASJC Scopus subject areas

  • Mathematics(all)
  • Statistics and Probability

Cite this

Spike and slab gene selection for multigroup microarray data. / Ishwaran, Hemant; Rao, Jonnagadda S.

In: Journal of the American Statistical Association, Vol. 100, No. 471, 01.09.2005, p. 764-780.

Research output: Contribution to journalArticle

@article{8f479057f94b4e178522a248b1d5fb75,
title = "Spike and slab gene selection for multigroup microarray data",
abstract = "DNA microarrays can provide insight into genetic changes that characterize different stages of a disease process. Accurate identification of these changes has significant therapeutic and diagnostic implications. Statistical analysis for multistage (multigroup) data is challenging, however. ANOVA-based extensions of two-sample Z-tests, a popular method for detecting differentially expressed genes in two groups, do not work well in multigroup settings. False detection rates are high because of variability of the ordinary least squares estimators and because of regression to the mean induced by correlated parameter estimates. We develop a Bayesian rescaled spike and slab hierarchical model specifically designed for the multigroup gene detection problem. Data preprocessing steps are introduced to deal with unique features of microarray data and to enhance selection performance. We show theoretically that spike and slab models naturally encourage sparse solutions through a process called selective shrinkage. This translates into oracle-like gene selection risk performance compared with ordinary least squares estimates. The methodology is illustrated on a large microarray repository of samples from different clinical stages of metastatic colon cancer. Through a functional analysis of selected genes, we show that spike and slab models identify important biological signals while minimizing biologically implausible false detections.",
keywords = "Colon cancer, Hypervariance, Penalization, Rescaling, Risk misclassification, Shrinkage, Sparsity, Stochastic variable selection, Zcut",
author = "Hemant Ishwaran and Rao, {Jonnagadda S}",
year = "2005",
month = "9",
day = "1",
doi = "10.1198/016214505000000051",
language = "English",
volume = "100",
pages = "764--780",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "471",

}

TY - JOUR

T1 - Spike and slab gene selection for multigroup microarray data

AU - Ishwaran, Hemant

AU - Rao, Jonnagadda S

PY - 2005/9/1

Y1 - 2005/9/1

N2 - DNA microarrays can provide insight into genetic changes that characterize different stages of a disease process. Accurate identification of these changes has significant therapeutic and diagnostic implications. Statistical analysis for multistage (multigroup) data is challenging, however. ANOVA-based extensions of two-sample Z-tests, a popular method for detecting differentially expressed genes in two groups, do not work well in multigroup settings. False detection rates are high because of variability of the ordinary least squares estimators and because of regression to the mean induced by correlated parameter estimates. We develop a Bayesian rescaled spike and slab hierarchical model specifically designed for the multigroup gene detection problem. Data preprocessing steps are introduced to deal with unique features of microarray data and to enhance selection performance. We show theoretically that spike and slab models naturally encourage sparse solutions through a process called selective shrinkage. This translates into oracle-like gene selection risk performance compared with ordinary least squares estimates. The methodology is illustrated on a large microarray repository of samples from different clinical stages of metastatic colon cancer. Through a functional analysis of selected genes, we show that spike and slab models identify important biological signals while minimizing biologically implausible false detections.

AB - DNA microarrays can provide insight into genetic changes that characterize different stages of a disease process. Accurate identification of these changes has significant therapeutic and diagnostic implications. Statistical analysis for multistage (multigroup) data is challenging, however. ANOVA-based extensions of two-sample Z-tests, a popular method for detecting differentially expressed genes in two groups, do not work well in multigroup settings. False detection rates are high because of variability of the ordinary least squares estimators and because of regression to the mean induced by correlated parameter estimates. We develop a Bayesian rescaled spike and slab hierarchical model specifically designed for the multigroup gene detection problem. Data preprocessing steps are introduced to deal with unique features of microarray data and to enhance selection performance. We show theoretically that spike and slab models naturally encourage sparse solutions through a process called selective shrinkage. This translates into oracle-like gene selection risk performance compared with ordinary least squares estimates. The methodology is illustrated on a large microarray repository of samples from different clinical stages of metastatic colon cancer. Through a functional analysis of selected genes, we show that spike and slab models identify important biological signals while minimizing biologically implausible false detections.

KW - Colon cancer

KW - Hypervariance

KW - Penalization

KW - Rescaling

KW - Risk misclassification

KW - Shrinkage

KW - Sparsity

KW - Stochastic variable selection

KW - Zcut

UR - http://www.scopus.com/inward/record.url?scp=22944431540&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=22944431540&partnerID=8YFLogxK

U2 - 10.1198/016214505000000051

DO - 10.1198/016214505000000051

M3 - Article

AN - SCOPUS:22944431540

VL - 100

SP - 764

EP - 780

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 471

ER -