Variable selection in Bayesian generalized linear-mixed models: An illustration using candidate gene case-control association studies

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

The problem of variable selection in the generalized linear-mixed models (GLMMs) is pervasive in statistical practice. For the purpose of variable selection, many methodologies for determining the best subset of explanatory variables currently exist according to the model complexity and differences between applications. In this paper, we develop a "higher posterior probability model with bootstrap" (HPMB) approach to select explanatory variables without fitting all possible GLMMs involving a small or moderate number of explanatory variables. Furthermore, to save computational load, we propose an efficient approximation approach with Laplace's method and Taylor's expansion to approximate intractable integrals in GLMMs. Simulation studies and an application of HapMap data provide evidence that this selection approach is computationally feasible and reliable for exploring true candidate genes and gene-gene associations, after adjusting for complex structures among clusters.

Original languageEnglish
Pages (from-to)234-253
Number of pages20
JournalBiometrical Journal
Volume57
Issue number2
DOIs
Publication statusPublished - 2015 Jan 1

Fingerprint

Generalized Linear Mixed Model
Case-control
Variable Selection
Case-Control Studies
Selection of Variables
Linear Models
Gene
HapMap Project
Genes
Laplace's Method
Model Complexity
Taylor Expansion
Probability Model
Posterior Probability
Complex Structure
Bootstrap
Simulation Study
Subset
Methodology
Approximation

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

@article{3923aae8742c4d5792d294c5cfe5d376,
title = "Variable selection in Bayesian generalized linear-mixed models: An illustration using candidate gene case-control association studies",
abstract = "The problem of variable selection in the generalized linear-mixed models (GLMMs) is pervasive in statistical practice. For the purpose of variable selection, many methodologies for determining the best subset of explanatory variables currently exist according to the model complexity and differences between applications. In this paper, we develop a {"}higher posterior probability model with bootstrap{"} (HPMB) approach to select explanatory variables without fitting all possible GLMMs involving a small or moderate number of explanatory variables. Furthermore, to save computational load, we propose an efficient approximation approach with Laplace's method and Taylor's expansion to approximate intractable integrals in GLMMs. Simulation studies and an application of HapMap data provide evidence that this selection approach is computationally feasible and reliable for exploring true candidate genes and gene-gene associations, after adjusting for complex structures among clusters.",
author = "Miao-Yu Tsai",
year = "2015",
month = "1",
day = "1",
doi = "10.1002/bimj.201300259",
language = "English",
volume = "57",
pages = "234--253",
journal = "Biometrical Journal",
issn = "0323-3847",
publisher = "Wiley-VCH Verlag",
number = "2",

}

TY - JOUR

T1 - Variable selection in Bayesian generalized linear-mixed models

T2 - An illustration using candidate gene case-control association studies

AU - Tsai, Miao-Yu

PY - 2015/1/1

Y1 - 2015/1/1

N2 - The problem of variable selection in the generalized linear-mixed models (GLMMs) is pervasive in statistical practice. For the purpose of variable selection, many methodologies for determining the best subset of explanatory variables currently exist according to the model complexity and differences between applications. In this paper, we develop a "higher posterior probability model with bootstrap" (HPMB) approach to select explanatory variables without fitting all possible GLMMs involving a small or moderate number of explanatory variables. Furthermore, to save computational load, we propose an efficient approximation approach with Laplace's method and Taylor's expansion to approximate intractable integrals in GLMMs. Simulation studies and an application of HapMap data provide evidence that this selection approach is computationally feasible and reliable for exploring true candidate genes and gene-gene associations, after adjusting for complex structures among clusters.

AB - The problem of variable selection in the generalized linear-mixed models (GLMMs) is pervasive in statistical practice. For the purpose of variable selection, many methodologies for determining the best subset of explanatory variables currently exist according to the model complexity and differences between applications. In this paper, we develop a "higher posterior probability model with bootstrap" (HPMB) approach to select explanatory variables without fitting all possible GLMMs involving a small or moderate number of explanatory variables. Furthermore, to save computational load, we propose an efficient approximation approach with Laplace's method and Taylor's expansion to approximate intractable integrals in GLMMs. Simulation studies and an application of HapMap data provide evidence that this selection approach is computationally feasible and reliable for exploring true candidate genes and gene-gene associations, after adjusting for complex structures among clusters.

UR - http://www.scopus.com/inward/record.url?scp=84923592466&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84923592466&partnerID=8YFLogxK

U2 - 10.1002/bimj.201300259

DO - 10.1002/bimj.201300259

M3 - Article

C2 - 25267186

AN - SCOPUS:84923592466

VL - 57

SP - 234

EP - 253

JO - Biometrical Journal

JF - Biometrical Journal

SN - 0323-3847

IS - 2

ER -