Bayesian longitudinal low-rank regression models for imaging genetic data from longitudinal studies

Zhao Hua Lu, Zakaria Khondker, Joseph G. Ibrahim, Yue Wang, Hongtu Zhu

Research output: Contribution to journalArticle

  • 1 Citations

Abstract

To perform a joint analysis of multivariate neuroimaging phenotypes and candidate genetic markers obtained from longitudinal studies, we develop a Bayesian longitudinal low-rank regression (L2R2) model. The L2R2 model integrates three key methodologies: a low-rank matrix for approximating the high-dimensional regression coefficient matrices corresponding to the genetic main effects and their interactions with time, penalized splines for characterizing the overall time effect, and a sparse factor analysis model coupled with random effects for capturing within-subject spatio-temporal correlations of longitudinal phenotypes. Posterior computation proceeds via an efficient Markov chain Monte Carlo algorithm. Simulations show that the L2R2 model outperforms several other competing methods. We apply the L2R2 model to investigate the effect of single nucleotide polymorphisms (SNPs) on the top 10 and top 40 previously reported Alzheimer disease-associated genes. We also identify associations between the interactions of these SNPs with patient age and the tissue volumes of 93 regions of interest from patients’ brain images obtained from the Alzheimer's Disease Neuroimaging Initiative.

LanguageEnglish (US)
Pages305-322
Number of pages18
JournalNeuroImage
Volume149
DOIs
StatePublished - Apr 1 2017

Fingerprint

Genetic Models
Longitudinal Studies
Neuroimaging
Single Nucleotide Polymorphism
Alzheimer Disease
Phenotype
Markov Chains
Genetic Markers
Statistical Factor Analysis
Multivariate Analysis
Brain
Genes

Keywords

  • Genetic variants
  • Longitudinal imaging phenotypes
  • Low-rank regression
  • Markov chain Monte Carlo
  • Spatiotemporal correlation

ASJC Scopus subject areas

  • Neurology
  • Cognitive Neuroscience

Cite this

Bayesian longitudinal low-rank regression models for imaging genetic data from longitudinal studies. / Lu, Zhao Hua; Khondker, Zakaria; Ibrahim, Joseph G.; Wang, Yue; Zhu, Hongtu.

In: NeuroImage, Vol. 149, 01.04.2017, p. 305-322.

Research output: Contribution to journalArticle

@article{d779d62f05844b1bb5be1b8f736de4be,
title = "Bayesian longitudinal low-rank regression models for imaging genetic data from longitudinal studies",
abstract = "To perform a joint analysis of multivariate neuroimaging phenotypes and candidate genetic markers obtained from longitudinal studies, we develop a Bayesian longitudinal low-rank regression (L2R2) model. The L2R2 model integrates three key methodologies: a low-rank matrix for approximating the high-dimensional regression coefficient matrices corresponding to the genetic main effects and their interactions with time, penalized splines for characterizing the overall time effect, and a sparse factor analysis model coupled with random effects for capturing within-subject spatio-temporal correlations of longitudinal phenotypes. Posterior computation proceeds via an efficient Markov chain Monte Carlo algorithm. Simulations show that the L2R2 model outperforms several other competing methods. We apply the L2R2 model to investigate the effect of single nucleotide polymorphisms (SNPs) on the top 10 and top 40 previously reported Alzheimer disease-associated genes. We also identify associations between the interactions of these SNPs with patient age and the tissue volumes of 93 regions of interest from patients’ brain images obtained from the Alzheimer's Disease Neuroimaging Initiative.",
keywords = "Genetic variants, Longitudinal imaging phenotypes, Low-rank regression, Markov chain Monte Carlo, Spatiotemporal correlation",
author = "Lu, {Zhao Hua} and Zakaria Khondker and Ibrahim, {Joseph G.} and Yue Wang and Hongtu Zhu",
year = "2017",
month = "4",
day = "1",
doi = "10.1016/j.neuroimage.2017.01.052",
language = "English (US)",
volume = "149",
pages = "305--322",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Bayesian longitudinal low-rank regression models for imaging genetic data from longitudinal studies

AU - Lu,Zhao Hua

AU - Khondker,Zakaria

AU - Ibrahim,Joseph G.

AU - Wang,Yue

AU - Zhu,Hongtu

PY - 2017/4/1

Y1 - 2017/4/1

N2 - To perform a joint analysis of multivariate neuroimaging phenotypes and candidate genetic markers obtained from longitudinal studies, we develop a Bayesian longitudinal low-rank regression (L2R2) model. The L2R2 model integrates three key methodologies: a low-rank matrix for approximating the high-dimensional regression coefficient matrices corresponding to the genetic main effects and their interactions with time, penalized splines for characterizing the overall time effect, and a sparse factor analysis model coupled with random effects for capturing within-subject spatio-temporal correlations of longitudinal phenotypes. Posterior computation proceeds via an efficient Markov chain Monte Carlo algorithm. Simulations show that the L2R2 model outperforms several other competing methods. We apply the L2R2 model to investigate the effect of single nucleotide polymorphisms (SNPs) on the top 10 and top 40 previously reported Alzheimer disease-associated genes. We also identify associations between the interactions of these SNPs with patient age and the tissue volumes of 93 regions of interest from patients’ brain images obtained from the Alzheimer's Disease Neuroimaging Initiative.

AB - To perform a joint analysis of multivariate neuroimaging phenotypes and candidate genetic markers obtained from longitudinal studies, we develop a Bayesian longitudinal low-rank regression (L2R2) model. The L2R2 model integrates three key methodologies: a low-rank matrix for approximating the high-dimensional regression coefficient matrices corresponding to the genetic main effects and their interactions with time, penalized splines for characterizing the overall time effect, and a sparse factor analysis model coupled with random effects for capturing within-subject spatio-temporal correlations of longitudinal phenotypes. Posterior computation proceeds via an efficient Markov chain Monte Carlo algorithm. Simulations show that the L2R2 model outperforms several other competing methods. We apply the L2R2 model to investigate the effect of single nucleotide polymorphisms (SNPs) on the top 10 and top 40 previously reported Alzheimer disease-associated genes. We also identify associations between the interactions of these SNPs with patient age and the tissue volumes of 93 regions of interest from patients’ brain images obtained from the Alzheimer's Disease Neuroimaging Initiative.

KW - Genetic variants

KW - Longitudinal imaging phenotypes

KW - Low-rank regression

KW - Markov chain Monte Carlo

KW - Spatiotemporal correlation

UR - http://www.scopus.com/inward/record.url?scp=85012188118&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85012188118&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2017.01.052

DO - 10.1016/j.neuroimage.2017.01.052

M3 - Article

VL - 149

SP - 305

EP - 322

JO - NeuroImage

T2 - NeuroImage

JF - NeuroImage

SN - 1053-8119

ER -