Contaminated normal modeling with application to microarray data analysis

Hongying Dai, Richard Charnigo

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

A contaminated beta model (1 - γ)B(1, 1) + γB(α, β) is often used to describe the distribution of P-values arising from a microarray experiment. The authors propose and examine a different approach: namely, using a contaminated normal model (1 - γ)N(0, σ2) + γN(μ, σ2) to describe the distribution of Z statistics or suitably transformed T statistics. The authors then address whether a researcher who has Z statistics should analyze them using the contaminated normal model or whether the Z statistics should be converted to P-values to be analyzed using the contaminated beta model. The authors also provide a decisiontheoretic perspective on the analysis of Z statistics.

Original languageEnglish (US)
Pages (from-to)315-332
Number of pages18
JournalCanadian Journal of Statistics
Volume38
Issue number3
DOIs
StatePublished - Sep 1 2010

Fingerprint

Microarray Data Analysis
Statistics
Modeling
Microarray
Model
Experiment

Keywords

  • Contaminated beta model
  • Contaminated normal model
  • D-test
  • MLRT
  • Maximum modified likelihood
  • Microarray
  • Mixture model
  • Omnibus test

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Contaminated normal modeling with application to microarray data analysis. / Dai, Hongying; Charnigo, Richard.

In: Canadian Journal of Statistics, Vol. 38, No. 3, 01.09.2010, p. 315-332.

Research output: Contribution to journalArticle

@article{b0086e4d45764e3dbe0108b4239549c6,
title = "Contaminated normal modeling with application to microarray data analysis",
abstract = "A contaminated beta model (1 - γ)B(1, 1) + γB(α, β) is often used to describe the distribution of P-values arising from a microarray experiment. The authors propose and examine a different approach: namely, using a contaminated normal model (1 - γ)N(0, σ2) + γN(μ, σ2) to describe the distribution of Z statistics or suitably transformed T statistics. The authors then address whether a researcher who has Z statistics should analyze them using the contaminated normal model or whether the Z statistics should be converted to P-values to be analyzed using the contaminated beta model. The authors also provide a decisiontheoretic perspective on the analysis of Z statistics.",
keywords = "Contaminated beta model, Contaminated normal model, D-test, MLRT, Maximum modified likelihood, Microarray, Mixture model, Omnibus test",
author = "Hongying Dai and Richard Charnigo",
year = "2010",
month = "9",
day = "1",
doi = "10.1002/cjs.10053",
language = "English (US)",
volume = "38",
pages = "315--332",
journal = "Canadian Journal of Statistics",
issn = "0319-5724",
publisher = "Statistical Society of Canada",
number = "3",

}

TY - JOUR

T1 - Contaminated normal modeling with application to microarray data analysis

AU - Dai, Hongying

AU - Charnigo, Richard

PY - 2010/9/1

Y1 - 2010/9/1

N2 - A contaminated beta model (1 - γ)B(1, 1) + γB(α, β) is often used to describe the distribution of P-values arising from a microarray experiment. The authors propose and examine a different approach: namely, using a contaminated normal model (1 - γ)N(0, σ2) + γN(μ, σ2) to describe the distribution of Z statistics or suitably transformed T statistics. The authors then address whether a researcher who has Z statistics should analyze them using the contaminated normal model or whether the Z statistics should be converted to P-values to be analyzed using the contaminated beta model. The authors also provide a decisiontheoretic perspective on the analysis of Z statistics.

AB - A contaminated beta model (1 - γ)B(1, 1) + γB(α, β) is often used to describe the distribution of P-values arising from a microarray experiment. The authors propose and examine a different approach: namely, using a contaminated normal model (1 - γ)N(0, σ2) + γN(μ, σ2) to describe the distribution of Z statistics or suitably transformed T statistics. The authors then address whether a researcher who has Z statistics should analyze them using the contaminated normal model or whether the Z statistics should be converted to P-values to be analyzed using the contaminated beta model. The authors also provide a decisiontheoretic perspective on the analysis of Z statistics.

KW - Contaminated beta model

KW - Contaminated normal model

KW - D-test

KW - MLRT

KW - Maximum modified likelihood

KW - Microarray

KW - Mixture model

KW - Omnibus test

UR - http://www.scopus.com/inward/record.url?scp=77956634780&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77956634780&partnerID=8YFLogxK

U2 - 10.1002/cjs.10053

DO - 10.1002/cjs.10053

M3 - Article

AN - SCOPUS:77956634780

VL - 38

SP - 315

EP - 332

JO - Canadian Journal of Statistics

JF - Canadian Journal of Statistics

SN - 0319-5724

IS - 3

ER -