Generalized finite mixture of multivariate regressions with applications to therapeutic biomarker identification

Hongmei Liu; J. Sunil Rao

doi:10.1002/sim.8726

Generalized finite mixture of multivariate regressions with applications to therapeutic biomarker identification

Research output: Contribution to journal › Article › peer-review

Abstract

Finite mixtures of regressions have been used to analyze data that come from a heterogeneous population. When more than one response is observed, accommodating a multivariate response can be useful. In this article, we go a step further and introduce a multivariate extension that includes a latent overlapping cluster indicator variable that allows for potential overdispersion. A generalized mixture of multivariate regressions in connection with the proposed model and a new EM algorithm for fitting are provided. In addition, we allow for high-dimensional predictors via shrinkage estimation. This model proves particularly useful in the analysis of complex data like the search for cancer therapeutic biomarkers. We demonstrate this using the genomics of drug sensitivity in cancer resource.

Original language	English (US)
Pages (from-to)	4301-4324
Number of pages	24
Journal	Statistics in Medicine
Volume	39
Issue number	28
DOIs	https://doi.org/10.1002/sim.8726
State	Published - Dec 10 2020
Externally published	Yes

Keywords

EM algorithm
Lasso
cancer biomarkers
mixture of multivariate regression models
overlapping clustering

ASJC Scopus subject areas

Epidemiology
Statistics and Probability

Access to Document

10.1002/sim.8726

Fingerprint Dive into the research topics of 'Generalized finite mixture of multivariate regressions with applications to therapeutic biomarker identification'. Together they form a unique fingerprint.

Cite this

@article{e680c195d5224c12a57f2d45bf55141d,

title = "Generalized finite mixture of multivariate regressions with applications to therapeutic biomarker identification",

abstract = "Finite mixtures of regressions have been used to analyze data that come from a heterogeneous population. When more than one response is observed, accommodating a multivariate response can be useful. In this article, we go a step further and introduce a multivariate extension that includes a latent overlapping cluster indicator variable that allows for potential overdispersion. A generalized mixture of multivariate regressions in connection with the proposed model and a new EM algorithm for fitting are provided. In addition, we allow for high-dimensional predictors via shrinkage estimation. This model proves particularly useful in the analysis of complex data like the search for cancer therapeutic biomarkers. We demonstrate this using the genomics of drug sensitivity in cancer resource.",

keywords = "EM algorithm, Lasso, cancer biomarkers, mixture of multivariate regression models, overlapping clustering",

author = "Hongmei Liu and {Sunil Rao}, J.",

note = "Funding Information: Hongmei Liu was supported by a doctoral fellowship from the Sylvester Comprehensive Cancer Center at the University of Miami. Publisher Copyright: {\textcopyright} 2020 John Wiley & Sons Ltd Copyright: Copyright 2021 Elsevier B.V., All rights reserved.",

year = "2020",

month = dec,

day = "10",

doi = "10.1002/sim.8726",

language = "English (US)",

volume = "39",

pages = "4301--4324",

journal = "Statistics in Medicine",

issn = "0277-6715",

publisher = "John Wiley and Sons Ltd",

number = "28",

}

TY - JOUR

T1 - Generalized finite mixture of multivariate regressions with applications to therapeutic biomarker identification

AU - Liu, Hongmei

AU - Sunil Rao, J.

N1 - Funding Information: Hongmei Liu was supported by a doctoral fellowship from the Sylvester Comprehensive Cancer Center at the University of Miami. Publisher Copyright: © 2020 John Wiley & Sons Ltd Copyright: Copyright 2021 Elsevier B.V., All rights reserved.

PY - 2020/12/10

Y1 - 2020/12/10

N2 - Finite mixtures of regressions have been used to analyze data that come from a heterogeneous population. When more than one response is observed, accommodating a multivariate response can be useful. In this article, we go a step further and introduce a multivariate extension that includes a latent overlapping cluster indicator variable that allows for potential overdispersion. A generalized mixture of multivariate regressions in connection with the proposed model and a new EM algorithm for fitting are provided. In addition, we allow for high-dimensional predictors via shrinkage estimation. This model proves particularly useful in the analysis of complex data like the search for cancer therapeutic biomarkers. We demonstrate this using the genomics of drug sensitivity in cancer resource.

AB - Finite mixtures of regressions have been used to analyze data that come from a heterogeneous population. When more than one response is observed, accommodating a multivariate response can be useful. In this article, we go a step further and introduce a multivariate extension that includes a latent overlapping cluster indicator variable that allows for potential overdispersion. A generalized mixture of multivariate regressions in connection with the proposed model and a new EM algorithm for fitting are provided. In addition, we allow for high-dimensional predictors via shrinkage estimation. This model proves particularly useful in the analysis of complex data like the search for cancer therapeutic biomarkers. We demonstrate this using the genomics of drug sensitivity in cancer resource.

KW - EM algorithm

KW - Lasso

KW - cancer biomarkers

KW - mixture of multivariate regression models

KW - overlapping clustering

UR - http://www.scopus.com/inward/record.url?scp=85089917041&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85089917041&partnerID=8YFLogxK

U2 - 10.1002/sim.8726

DO - 10.1002/sim.8726

M3 - Article

C2 - 32857877

AN - SCOPUS:85089917041

VL - 39

SP - 4301

EP - 4324

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 28

ER -