Prior robust empirical Bayes inference for large-scale data by conditioning on rank with application to microarray data

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Empirical Bayes methods have been extensively used for microarray data analysis by modeling the large number of unknown parameters as random effects. Empirical Bayes allows borrowing information across genes and can automatically adjust for multiple testing and selection bias. However, the standard empirical Bayes model can perform poorly if the assumed working prior deviates from the true prior. This paper proposes a new rank-conditioned inference in which the shrinkage and confidence intervals are based on the distribution of the error conditioned on rank of the data. Our approach is in contrast to a Bayesian posterior, which conditions on the data themselves. The new method is almost as efficient as standard Bayesian methods when the working prior is close to the true prior, and it is much more robust when the working prior is not close. In addition, it allows a more accurate (but also more complex) non-parametric estimate of the prior to be easily incorporated, resulting in improved inference. The new method's prior robustness is demonstrated via simulation experiments. Application to a breast cancer gene expression microarray dataset is presented. Our R package rank. Shrinkage provides a ready-to-use implementation of the proposed methodology.

Original languageEnglish (US)
Pages (from-to)63-73
Number of pages11
JournalBiostatistics
Volume15
Issue number1
DOIs
StatePublished - Jan 1 2014

Fingerprint

Empirical Bayes
Microarray Data
Conditioning
Shrinkage
Empirical Bayes Method
Microarray Data Analysis
Selection Bias
Multiple Testing
Bayesian Methods
Random Effects
Breast Cancer
Microarray
Unknown Parameters
Simulation Experiment
Gene Expression
Confidence interval
Robustness
Gene
Methodology
Modeling

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

@article{4ac058cf025d4c1bb183cd32481a21db,
title = "Prior robust empirical Bayes inference for large-scale data by conditioning on rank with application to microarray data",
abstract = "Empirical Bayes methods have been extensively used for microarray data analysis by modeling the large number of unknown parameters as random effects. Empirical Bayes allows borrowing information across genes and can automatically adjust for multiple testing and selection bias. However, the standard empirical Bayes model can perform poorly if the assumed working prior deviates from the true prior. This paper proposes a new rank-conditioned inference in which the shrinkage and confidence intervals are based on the distribution of the error conditioned on rank of the data. Our approach is in contrast to a Bayesian posterior, which conditions on the data themselves. The new method is almost as efficient as standard Bayesian methods when the working prior is close to the true prior, and it is much more robust when the working prior is not close. In addition, it allows a more accurate (but also more complex) non-parametric estimate of the prior to be easily incorporated, resulting in improved inference. The new method's prior robustness is demonstrated via simulation experiments. Application to a breast cancer gene expression microarray dataset is presented. Our R package rank. Shrinkage provides a ready-to-use implementation of the proposed methodology.",
author = "Liao, {Jiangang (Jason)} and Timothy Mcmurry and Arthur Berg",
year = "2014",
month = "1",
day = "1",
doi = "10.1093/biostatistics/kxt026",
language = "English (US)",
volume = "15",
pages = "63--73",
journal = "Biostatistics",
issn = "1465-4644",
publisher = "Oxford University Press",
number = "1",

}

Prior robust empirical Bayes inference for large-scale data by conditioning on rank with application to microarray data. / Liao, Jiangang (Jason); Mcmurry, Timothy; Berg, Arthur.

In: Biostatistics, Vol. 15, No. 1, 01.01.2014, p. 63-73.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Prior robust empirical Bayes inference for large-scale data by conditioning on rank with application to microarray data

AU - Liao, Jiangang (Jason)

AU - Mcmurry, Timothy

AU - Berg, Arthur

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Empirical Bayes methods have been extensively used for microarray data analysis by modeling the large number of unknown parameters as random effects. Empirical Bayes allows borrowing information across genes and can automatically adjust for multiple testing and selection bias. However, the standard empirical Bayes model can perform poorly if the assumed working prior deviates from the true prior. This paper proposes a new rank-conditioned inference in which the shrinkage and confidence intervals are based on the distribution of the error conditioned on rank of the data. Our approach is in contrast to a Bayesian posterior, which conditions on the data themselves. The new method is almost as efficient as standard Bayesian methods when the working prior is close to the true prior, and it is much more robust when the working prior is not close. In addition, it allows a more accurate (but also more complex) non-parametric estimate of the prior to be easily incorporated, resulting in improved inference. The new method's prior robustness is demonstrated via simulation experiments. Application to a breast cancer gene expression microarray dataset is presented. Our R package rank. Shrinkage provides a ready-to-use implementation of the proposed methodology.

AB - Empirical Bayes methods have been extensively used for microarray data analysis by modeling the large number of unknown parameters as random effects. Empirical Bayes allows borrowing information across genes and can automatically adjust for multiple testing and selection bias. However, the standard empirical Bayes model can perform poorly if the assumed working prior deviates from the true prior. This paper proposes a new rank-conditioned inference in which the shrinkage and confidence intervals are based on the distribution of the error conditioned on rank of the data. Our approach is in contrast to a Bayesian posterior, which conditions on the data themselves. The new method is almost as efficient as standard Bayesian methods when the working prior is close to the true prior, and it is much more robust when the working prior is not close. In addition, it allows a more accurate (but also more complex) non-parametric estimate of the prior to be easily incorporated, resulting in improved inference. The new method's prior robustness is demonstrated via simulation experiments. Application to a breast cancer gene expression microarray dataset is presented. Our R package rank. Shrinkage provides a ready-to-use implementation of the proposed methodology.

UR - http://www.scopus.com/inward/record.url?scp=84890537645&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890537645&partnerID=8YFLogxK

U2 - 10.1093/biostatistics/kxt026

DO - 10.1093/biostatistics/kxt026

M3 - Article

VL - 15

SP - 63

EP - 73

JO - Biostatistics

JF - Biostatistics

SN - 1465-4644

IS - 1

ER -