Sparse time series chain graphical models for reconstructing genetic networks

Fentaw Abegaz*, Ernst Wit

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

74 Citations (Scopus)

Abstract

We propose a sparse high-dimensional time series chain graphical model for reconstructing genetic networks from gene expression data parametrized by a precision matrix and autoregressive coefficient matrix. We consider the time steps as blocks or chains. The proposed approach explores patterns of contemporaneous and dynamic interactions by efficiently combining Gaussian graphical models and Bayesian dynamic networks. We use penalized likelihood inference with a smoothly clipped absolute deviation penalty to explore the relationships among the observed time course gene expressions. The method is illustrated on simulated data and on real data examples from Arabidopsis thaliana and mammary gland time course microarray gene expressions.

Original languageEnglish
Pages (from-to)586-599
Number of pages14
JournalBiostatistics
Volume14
Issue number3
DOIs
Publication statusPublished - Jul-2013

Keywords

  • Chain graphical mode
  • Dynamic network
  • Gene expression
  • High-dimensional data
  • L-1 penalty
  • Model selection
  • Penalized likelihood
  • SCAD penalty
  • Vector autoregressive model
  • NONCONCAVE PENALIZED LIKELIHOOD
  • COVARIANCE ESTIMATION
  • ARABIDOPSIS-THALIANA
  • VARIABLE SELECTION
  • MATRIX ESTIMATION
  • MAMMARY-GLAND
  • LASSO
  • REGRESSION
  • BIOLOGY
  • CLOCK

Fingerprint

Dive into the research topics of 'Sparse time series chain graphical models for reconstructing genetic networks'. Together they form a unique fingerprint.

Cite this