Inferring gene regulatory networks via nonlinear state-space models and exploiting sparsity

Research output: Contribution to journalArticle

29 Citations (Scopus)

Abstract

This paper considers the problem of learning the structure of gene regulatory networks from gene expression time series data. A more realistic scenario when the state space model representing a gene network evolves nonlinearly is considered while a linear model is assumed for the microarray data. To capture the nonlinearity, a particle filter-based state estimation algorithm is considered instead of the contemporary linear approximation-based approaches. The parameters characterizing the regulatory relations among various genes are estimated online using a Kalman filter. Since a particular gene interacts with a few other genes only, the parameter vector is expected to be sparse. The state estimates delivered by the particle filter and the observed microarray data are then subjected to a LASSO-based least squares regression operation which yields a parsimonious and efficient description of the regulatory network by setting the irrelevant coefficients to zero. The performance of the aforementioned algorithm is compared with the extended Kalman filter (EKF) and Unscented Kalman Filter (UKF) employing the Mean Square Error (MSE) as the fidelity criterion in recovering the parameters of gene regulatory networks from synthetic data and real biological data. Extensive computer simulations illustrate that the proposed particle filter-based network inference algorithm outperforms EKF and UKF, and therefore, it can serve as a natural framework for modeling gene regulatory networks with nonlinear and sparse structure.

Original languageEnglish
Article number6202798
Pages (from-to)1203-1211
Number of pages9
JournalIEEE/ACM Transactions on Computational Biology and Bioinformatics
Volume9
Issue number4
DOIs
Publication statusPublished - 2012

Fingerprint

Space Simulation
Gene Regulatory Networks
Gene Regulatory Network
State-space Model
Sparsity
Kalman Filter
Nonlinear Model
Genes
Particle Filter
Kalman filters
Microarray Data
Gene
Extended Kalman filters
Microarrays
Regulator Genes
Least-Squares Analysis
Gene Networks
Least Squares Regression
Computer Simulation
Regulatory Networks

Keywords

  • Gene regulatory network
  • Kalman filter
  • LASSO
  • parameter estimation
  • particle filter

ASJC Scopus subject areas

  • Biotechnology
  • Genetics
  • Applied Mathematics

Cite this

@article{c9793d12cd40405ba7ecefaee1d1de80,
title = "Inferring gene regulatory networks via nonlinear state-space models and exploiting sparsity",
abstract = "This paper considers the problem of learning the structure of gene regulatory networks from gene expression time series data. A more realistic scenario when the state space model representing a gene network evolves nonlinearly is considered while a linear model is assumed for the microarray data. To capture the nonlinearity, a particle filter-based state estimation algorithm is considered instead of the contemporary linear approximation-based approaches. The parameters characterizing the regulatory relations among various genes are estimated online using a Kalman filter. Since a particular gene interacts with a few other genes only, the parameter vector is expected to be sparse. The state estimates delivered by the particle filter and the observed microarray data are then subjected to a LASSO-based least squares regression operation which yields a parsimonious and efficient description of the regulatory network by setting the irrelevant coefficients to zero. The performance of the aforementioned algorithm is compared with the extended Kalman filter (EKF) and Unscented Kalman Filter (UKF) employing the Mean Square Error (MSE) as the fidelity criterion in recovering the parameters of gene regulatory networks from synthetic data and real biological data. Extensive computer simulations illustrate that the proposed particle filter-based network inference algorithm outperforms EKF and UKF, and therefore, it can serve as a natural framework for modeling gene regulatory networks with nonlinear and sparse structure.",
keywords = "Gene regulatory network, Kalman filter, LASSO, parameter estimation, particle filter",
author = "Amina Noor and Erchin Serpedin and Mohamed Nounou and Hazem Nounou",
year = "2012",
doi = "10.1109/TCBB.2012.32",
language = "English",
volume = "9",
pages = "1203--1211",
journal = "IEEE/ACM Transactions on Computational Biology and Bioinformatics",
issn = "1545-5963",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Inferring gene regulatory networks via nonlinear state-space models and exploiting sparsity

AU - Noor, Amina

AU - Serpedin, Erchin

AU - Nounou, Mohamed

AU - Nounou, Hazem

PY - 2012

Y1 - 2012

N2 - This paper considers the problem of learning the structure of gene regulatory networks from gene expression time series data. A more realistic scenario when the state space model representing a gene network evolves nonlinearly is considered while a linear model is assumed for the microarray data. To capture the nonlinearity, a particle filter-based state estimation algorithm is considered instead of the contemporary linear approximation-based approaches. The parameters characterizing the regulatory relations among various genes are estimated online using a Kalman filter. Since a particular gene interacts with a few other genes only, the parameter vector is expected to be sparse. The state estimates delivered by the particle filter and the observed microarray data are then subjected to a LASSO-based least squares regression operation which yields a parsimonious and efficient description of the regulatory network by setting the irrelevant coefficients to zero. The performance of the aforementioned algorithm is compared with the extended Kalman filter (EKF) and Unscented Kalman Filter (UKF) employing the Mean Square Error (MSE) as the fidelity criterion in recovering the parameters of gene regulatory networks from synthetic data and real biological data. Extensive computer simulations illustrate that the proposed particle filter-based network inference algorithm outperforms EKF and UKF, and therefore, it can serve as a natural framework for modeling gene regulatory networks with nonlinear and sparse structure.

AB - This paper considers the problem of learning the structure of gene regulatory networks from gene expression time series data. A more realistic scenario when the state space model representing a gene network evolves nonlinearly is considered while a linear model is assumed for the microarray data. To capture the nonlinearity, a particle filter-based state estimation algorithm is considered instead of the contemporary linear approximation-based approaches. The parameters characterizing the regulatory relations among various genes are estimated online using a Kalman filter. Since a particular gene interacts with a few other genes only, the parameter vector is expected to be sparse. The state estimates delivered by the particle filter and the observed microarray data are then subjected to a LASSO-based least squares regression operation which yields a parsimonious and efficient description of the regulatory network by setting the irrelevant coefficients to zero. The performance of the aforementioned algorithm is compared with the extended Kalman filter (EKF) and Unscented Kalman Filter (UKF) employing the Mean Square Error (MSE) as the fidelity criterion in recovering the parameters of gene regulatory networks from synthetic data and real biological data. Extensive computer simulations illustrate that the proposed particle filter-based network inference algorithm outperforms EKF and UKF, and therefore, it can serve as a natural framework for modeling gene regulatory networks with nonlinear and sparse structure.

KW - Gene regulatory network

KW - Kalman filter

KW - LASSO

KW - parameter estimation

KW - particle filter

UR - http://www.scopus.com/inward/record.url?scp=84861507910&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84861507910&partnerID=8YFLogxK

U2 - 10.1109/TCBB.2012.32

DO - 10.1109/TCBB.2012.32

M3 - Article

VL - 9

SP - 1203

EP - 1211

JO - IEEE/ACM Transactions on Computational Biology and Bioinformatics

JF - IEEE/ACM Transactions on Computational Biology and Bioinformatics

SN - 1545-5963

IS - 4

M1 - 6202798

ER -