Estimation of multiple networks in Gaussian mixture models

Chen Gao, Yunzhang Zhu, Xiaotong T Shen, Wei Pan

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

We aim to estimate multiple networks in the presence of sample heterogeneity, where the independent samples (i.e. observations) may come from different and unknown populations or distributions. Specifically, we consider penalized estimation of multiple precision matrices in the framework of a Gaussian mixture model. A major innovation is to take advantage of the commonalities across the multiple precision matrices through possibly nonconvex fusion regularization, which for example makes it possible to achieve simultaneous discovery of unknown disease subtypes and detection of differential gene (dys)regulations in functional genomics. We embed in the EM algorithm one of two recently proposed methods for estimating multiple precision matrices in Gaussian graphical models. We demonstrate the feasibility and potential usefulness of the proposed methods in an application to glioblastoma subtype discovery and differential gene network analysis with a microarray gene expression data set. We also conduct realistic simulation studies to evaluate and compare the performance of various methods.

Original languageEnglish (US)
Pages (from-to)1133-1154
Number of pages22
JournalElectronic Journal of Statistics
Volume10
Issue number1
DOIs
StatePublished - Jan 1 2016

Fingerprint

Gaussian Mixture Model
Functional Genomics
Unknown
Gene Networks
Gene Regulation
Network Analysis
Gaussian Model
Graphical Models
EM Algorithm
Gene Expression Data
Microarray Data
Regularization
Fusion
Simulation Study
Evaluate
Estimate
Demonstrate
Gaussian mixture model
Gene

Keywords

  • Disease subtype discovery
  • Gaussian graphical model
  • Gene expression
  • Glioblastoma
  • Model-based clustering
  • Non-convex penalty

Cite this

Estimation of multiple networks in Gaussian mixture models. / Gao, Chen; Zhu, Yunzhang; Shen, Xiaotong T; Pan, Wei.

In: Electronic Journal of Statistics, Vol. 10, No. 1, 01.01.2016, p. 1133-1154.

Research output: Contribution to journalArticle

@article{0838ce2a2093421a860c52a54abd3938,
title = "Estimation of multiple networks in Gaussian mixture models",
abstract = "We aim to estimate multiple networks in the presence of sample heterogeneity, where the independent samples (i.e. observations) may come from different and unknown populations or distributions. Specifically, we consider penalized estimation of multiple precision matrices in the framework of a Gaussian mixture model. A major innovation is to take advantage of the commonalities across the multiple precision matrices through possibly nonconvex fusion regularization, which for example makes it possible to achieve simultaneous discovery of unknown disease subtypes and detection of differential gene (dys)regulations in functional genomics. We embed in the EM algorithm one of two recently proposed methods for estimating multiple precision matrices in Gaussian graphical models. We demonstrate the feasibility and potential usefulness of the proposed methods in an application to glioblastoma subtype discovery and differential gene network analysis with a microarray gene expression data set. We also conduct realistic simulation studies to evaluate and compare the performance of various methods.",
keywords = "Disease subtype discovery, Gaussian graphical model, Gene expression, Glioblastoma, Model-based clustering, Non-convex penalty",
author = "Chen Gao and Yunzhang Zhu and Shen, {Xiaotong T} and Wei Pan",
year = "2016",
month = "1",
day = "1",
doi = "10.1214/16-EJS1135",
language = "English (US)",
volume = "10",
pages = "1133--1154",
journal = "Electronic Journal of Statistics",
issn = "1935-7524",
publisher = "Institute of Mathematical Statistics",
number = "1",

}

TY - JOUR

T1 - Estimation of multiple networks in Gaussian mixture models

AU - Gao, Chen

AU - Zhu, Yunzhang

AU - Shen, Xiaotong T

AU - Pan, Wei

PY - 2016/1/1

Y1 - 2016/1/1

N2 - We aim to estimate multiple networks in the presence of sample heterogeneity, where the independent samples (i.e. observations) may come from different and unknown populations or distributions. Specifically, we consider penalized estimation of multiple precision matrices in the framework of a Gaussian mixture model. A major innovation is to take advantage of the commonalities across the multiple precision matrices through possibly nonconvex fusion regularization, which for example makes it possible to achieve simultaneous discovery of unknown disease subtypes and detection of differential gene (dys)regulations in functional genomics. We embed in the EM algorithm one of two recently proposed methods for estimating multiple precision matrices in Gaussian graphical models. We demonstrate the feasibility and potential usefulness of the proposed methods in an application to glioblastoma subtype discovery and differential gene network analysis with a microarray gene expression data set. We also conduct realistic simulation studies to evaluate and compare the performance of various methods.

AB - We aim to estimate multiple networks in the presence of sample heterogeneity, where the independent samples (i.e. observations) may come from different and unknown populations or distributions. Specifically, we consider penalized estimation of multiple precision matrices in the framework of a Gaussian mixture model. A major innovation is to take advantage of the commonalities across the multiple precision matrices through possibly nonconvex fusion regularization, which for example makes it possible to achieve simultaneous discovery of unknown disease subtypes and detection of differential gene (dys)regulations in functional genomics. We embed in the EM algorithm one of two recently proposed methods for estimating multiple precision matrices in Gaussian graphical models. We demonstrate the feasibility and potential usefulness of the proposed methods in an application to glioblastoma subtype discovery and differential gene network analysis with a microarray gene expression data set. We also conduct realistic simulation studies to evaluate and compare the performance of various methods.

KW - Disease subtype discovery

KW - Gaussian graphical model

KW - Gene expression

KW - Glioblastoma

KW - Model-based clustering

KW - Non-convex penalty

UR - http://www.scopus.com/inward/record.url?scp=84969242600&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84969242600&partnerID=8YFLogxK

U2 - 10.1214/16-EJS1135

DO - 10.1214/16-EJS1135

M3 - Article

C2 - 28966702

AN - SCOPUS:84969242600

VL - 10

SP - 1133

EP - 1154

JO - Electronic Journal of Statistics

JF - Electronic Journal of Statistics

SN - 1935-7524

IS - 1

ER -