Fused lasso regression for identifying differential correlations in brain connectome graphs

Donghyeon Yu; Sang Han Lee; Johan Lim; Guanghua Xiao; Richard Cameron Craddock; Bharat B. Biswal

doi:10.1002/sam.11382

Fused lasso regression for identifying differential correlations in brain connectome graphs

Donghyeon Yu, Sang Han Lee, Johan Lim, Guanghua Xiao, Richard Cameron Craddock, Bharat B. Biswal

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

3 Scopus citations

Abstract

In this paper, we propose a procedure to find differential edges between 2 graphs from high-dimensional data. We estimate 2 matrices of partial correlations and their differences by solving a penalized regression problem. We assume sparsity only on differences between 2 graphs, not graphs themselves. Thus, we impose an ℓ₂ penalty on partial correlations and an ℓ₁ penalty on their differences in the penalized regression problem. We apply the proposed procedure in finding differential functional connectivity between healthy individuals and Alzheimer's disease patients.

Original language	English (US)
Pages (from-to)	203-226
Number of pages	24
Journal	Statistical Analysis and Data Mining
Volume	11
Issue number	5
DOIs	https://doi.org/10.1002/sam.11382
State	Published - Oct 2018

All Science Journal Classification (ASJC) codes

Analysis
Information Systems
Computer Science Applications

Keywords

Gaussian graphical model
fMRI
functional connectivity
fusion penalty
partial correlation
penalized least squares
precision matrix

Access to Document

10.1002/sam.11382

Cite this

@article{457ee6fd96004e6d9ca604cd735f4b56,

title = "Fused lasso regression for identifying differential correlations in brain connectome graphs",

abstract = "In this paper, we propose a procedure to find differential edges between 2 graphs from high-dimensional data. We estimate 2 matrices of partial correlations and their differences by solving a penalized regression problem. We assume sparsity only on differences between 2 graphs, not graphs themselves. Thus, we impose an ℓ2 penalty on partial correlations and an ℓ1 penalty on their differences in the penalized regression problem. We apply the proposed procedure in finding differential functional connectivity between healthy individuals and Alzheimer's disease patients.",

keywords = "Gaussian graphical model, fMRI, functional connectivity, fusion penalty, partial correlation, penalized least squares, precision matrix",

author = "Donghyeon Yu and Lee, {Sang Han} and Johan Lim and Guanghua Xiao and Craddock, {Richard Cameron} and Biswal, {Bharat B.}",

note = "Publisher Copyright: {\textcopyright} 2018 Wiley Periodicals, Inc.",

year = "2018",

month = oct,

doi = "10.1002/sam.11382",

language = "English (US)",

volume = "11",

pages = "203--226",

journal = "Statistical Analysis and Data Mining",

issn = "1932-1864",

publisher = "John Wiley and Sons Inc.",

number = "5",

}

TY - JOUR

T1 - Fused lasso regression for identifying differential correlations in brain connectome graphs

AU - Yu, Donghyeon

AU - Lee, Sang Han

AU - Lim, Johan

AU - Xiao, Guanghua

AU - Craddock, Richard Cameron

AU - Biswal, Bharat B.

PY - 2018/10

Y1 - 2018/10

N2 - In this paper, we propose a procedure to find differential edges between 2 graphs from high-dimensional data. We estimate 2 matrices of partial correlations and their differences by solving a penalized regression problem. We assume sparsity only on differences between 2 graphs, not graphs themselves. Thus, we impose an ℓ2 penalty on partial correlations and an ℓ1 penalty on their differences in the penalized regression problem. We apply the proposed procedure in finding differential functional connectivity between healthy individuals and Alzheimer's disease patients.

AB - In this paper, we propose a procedure to find differential edges between 2 graphs from high-dimensional data. We estimate 2 matrices of partial correlations and their differences by solving a penalized regression problem. We assume sparsity only on differences between 2 graphs, not graphs themselves. Thus, we impose an ℓ2 penalty on partial correlations and an ℓ1 penalty on their differences in the penalized regression problem. We apply the proposed procedure in finding differential functional connectivity between healthy individuals and Alzheimer's disease patients.

KW - Gaussian graphical model

KW - fMRI

KW - functional connectivity

KW - fusion penalty

KW - partial correlation

KW - penalized least squares

KW - precision matrix

UR - http://www.scopus.com/inward/record.url?scp=85050933016&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85050933016&partnerID=8YFLogxK

U2 - 10.1002/sam.11382

DO - 10.1002/sam.11382

M3 - Article

AN - SCOPUS:85050933016

SN - 1932-1864

VL - 11

SP - 203

EP - 226

JO - Statistical Analysis and Data Mining

JF - Statistical Analysis and Data Mining

IS - 5

ER -

Fused lasso regression for identifying differential correlations in brain connectome graphs

Abstract

All Science Journal Classification (ASJC) codes

Keywords

Access to Document

Other files and links

Fingerprint

Cite this