Estimation of joint directed acyclic graphs with lasso family for gene networks

Sung Won Han; Sunghoon Park; Hua Zhong; Eun Seok Ryu; Pei Wang; Sehee Jung; Jayeon Lim; Jeewhan Yoon; Sung Hwan Kim

doi:10.1080/03610918.2019.1618869

Estimation of joint directed acyclic graphs with lasso family for gene networks

Sung Won Han, Sunghoon Park, Hua Zhong, Eun Seok Ryu, Pei Wang, Sehee Jung, Jayeon Lim, Jeewhan Yoon, Sung Hwan Kim

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

Abstract

Biological regulatory pathways provide important information for target gene cancer therapy. Frequently, estimating the gene networks of two distinct patient groups is a worthwhile investigation. This paper proposes an approach, called jDAG, to the estimation of directed joint networks. It can identify common directed edges with joint data sets and distinct edges. In a simulation study, we show that the proposed jDAG outperforms existing methods although it does require longer computational times. We also present and discuss the example study of a breast cancer data set with ER + and ER-.

Original language	English
Journal	Communications in Statistics: Simulation and Computation
DOIs	https://doi.org/10.1080/03610918.2019.1618869
Publication status	Published - 2019 Jan 1

Keywords

Bayesian network
Drug response network
Lasso estimation
Probabilistic graphical model
Structure equation model
Unknown natural ordering

ASJC Scopus subject areas

Statistics and Probability
Modelling and Simulation

Access to Document

10.1080/03610918.2019.1618869

Fingerprint Dive into the research topics of 'Estimation of joint directed acyclic graphs with lasso family for gene networks'. Together they form a unique fingerprint.

Cite this

@article{ab433d91124c4c51b95abb85b65b646c,

title = "Estimation of joint directed acyclic graphs with lasso family for gene networks",

abstract = "Biological regulatory pathways provide important information for target gene cancer therapy. Frequently, estimating the gene networks of two distinct patient groups is a worthwhile investigation. This paper proposes an approach, called jDAG, to the estimation of directed joint networks. It can identify common directed edges with joint data sets and distinct edges. In a simulation study, we show that the proposed jDAG outperforms existing methods although it does require longer computational times. We also present and discuss the example study of a breast cancer data set with ER + and ER-.",

keywords = "Bayesian network, Drug response network, Lasso estimation, Probabilistic graphical model, Structure equation model, Unknown natural ordering",

author = "Han, {Sung Won} and Sunghoon Park and Hua Zhong and Ryu, {Eun Seok} and Pei Wang and Sehee Jung and Jayeon Lim and Jeewhan Yoon and Kim, {Sung Hwan}",

year = "2019",

month = jan,

day = "1",

doi = "10.1080/03610918.2019.1618869",

language = "English",

journal = "Communications in Statistics Part B: Simulation and Computation",

issn = "0361-0918",

publisher = "Taylor and Francis Ltd.",

}

TY - JOUR

T1 - Estimation of joint directed acyclic graphs with lasso family for gene networks

AU - Han, Sung Won

AU - Park, Sunghoon

AU - Zhong, Hua

AU - Ryu, Eun Seok

AU - Wang, Pei

AU - Jung, Sehee

AU - Lim, Jayeon

AU - Yoon, Jeewhan

AU - Kim, Sung Hwan

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Biological regulatory pathways provide important information for target gene cancer therapy. Frequently, estimating the gene networks of two distinct patient groups is a worthwhile investigation. This paper proposes an approach, called jDAG, to the estimation of directed joint networks. It can identify common directed edges with joint data sets and distinct edges. In a simulation study, we show that the proposed jDAG outperforms existing methods although it does require longer computational times. We also present and discuss the example study of a breast cancer data set with ER + and ER-.

AB - Biological regulatory pathways provide important information for target gene cancer therapy. Frequently, estimating the gene networks of two distinct patient groups is a worthwhile investigation. This paper proposes an approach, called jDAG, to the estimation of directed joint networks. It can identify common directed edges with joint data sets and distinct edges. In a simulation study, we show that the proposed jDAG outperforms existing methods although it does require longer computational times. We also present and discuss the example study of a breast cancer data set with ER + and ER-.

KW - Bayesian network

KW - Drug response network

KW - Lasso estimation

KW - Probabilistic graphical model

KW - Structure equation model

KW - Unknown natural ordering

UR - http://www.scopus.com/inward/record.url?scp=85067571934&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067571934&partnerID=8YFLogxK

U2 - 10.1080/03610918.2019.1618869

DO - 10.1080/03610918.2019.1618869

M3 - Article

AN - SCOPUS:85067571934

JO - Communications in Statistics Part B: Simulation and Computation

JF - Communications in Statistics Part B: Simulation and Computation

SN - 0361-0918

ER -