Reverse engineering gene networks using global-local shrinkage rules

Viral Panchal, Daniel F. Linder

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Inferring gene regulatory networks from high-throughput 'omics' data has proven to be a computationally demanding task of critical importance. Frequently, the classical methods break down owing to the curse of dimensionality, and popular strategies to overcome this are typically based on regularized versions of the classical methods. However, these approaches rely on loss functions that may not be robust and usually do not allow for the incorporation of prior information in a straightforward way. Fully Bayesian methods are equipped to handle both of these shortcomings quite naturally, and they offer the potential for improvements in network structure learning. We propose a Bayesian hierarchical model to reconstruct gene regulatory networks from time-series gene expression data, such as those common in perturbation experiments of biological systems. The proposed methodology uses global-local shrinkage priors for posterior selection of regulatory edges and relaxes the common normal likelihood assumption in order to allow for heavy-tailed data, which were shown in several of the cited references to severely impact network inference. We provide a sufficient condition for posterior propriety and derive an efficient Markov chain Monte Carlo via Gibbs sampling in the electronic supplementary material. We describe a novel way to detect multiple scales based on the corresponding posterior quantities. Finally, we demonstrate the performance of our approach in a simulation study and compare it with existing methods on real data from a T-cell activation study.

Original languageEnglish (US)
Article number20190049
JournalInterface Focus
Volume10
Issue number1
DOIs
StatePublished - Feb 6 2020

Fingerprint

Reverse engineering
Gene Regulatory Networks
Genes
T-cells
Biological systems
Gene expression
Markov processes
Time series
Markov Chains
Bayes Theorem
Chemical activation
Throughput
Sampling
Learning
T-Lymphocytes
Gene Expression
Experiments

Keywords

  • Bayesian shrinkage
  • Gene networks
  • Horseshoe prior
  • Reverse engineering

ASJC Scopus subject areas

  • Biotechnology
  • Biophysics
  • Bioengineering
  • Biochemistry
  • Biomaterials
  • Biomedical Engineering

Cite this

Reverse engineering gene networks using global-local shrinkage rules. / Panchal, Viral; Linder, Daniel F.

In: Interface Focus, Vol. 10, No. 1, 20190049, 06.02.2020.

Research output: Contribution to journalArticle

@article{183b50911bfd4140a78f418e9dbf272f,
title = "Reverse engineering gene networks using global-local shrinkage rules",
abstract = "Inferring gene regulatory networks from high-throughput 'omics' data has proven to be a computationally demanding task of critical importance. Frequently, the classical methods break down owing to the curse of dimensionality, and popular strategies to overcome this are typically based on regularized versions of the classical methods. However, these approaches rely on loss functions that may not be robust and usually do not allow for the incorporation of prior information in a straightforward way. Fully Bayesian methods are equipped to handle both of these shortcomings quite naturally, and they offer the potential for improvements in network structure learning. We propose a Bayesian hierarchical model to reconstruct gene regulatory networks from time-series gene expression data, such as those common in perturbation experiments of biological systems. The proposed methodology uses global-local shrinkage priors for posterior selection of regulatory edges and relaxes the common normal likelihood assumption in order to allow for heavy-tailed data, which were shown in several of the cited references to severely impact network inference. We provide a sufficient condition for posterior propriety and derive an efficient Markov chain Monte Carlo via Gibbs sampling in the electronic supplementary material. We describe a novel way to detect multiple scales based on the corresponding posterior quantities. Finally, we demonstrate the performance of our approach in a simulation study and compare it with existing methods on real data from a T-cell activation study.",
keywords = "Bayesian shrinkage, Gene networks, Horseshoe prior, Reverse engineering",
author = "Viral Panchal and Linder, {Daniel F.}",
year = "2020",
month = "2",
day = "6",
doi = "10.1098/rsfs.2019.0049",
language = "English (US)",
volume = "10",
journal = "Interface Focus",
issn = "2042-8898",
publisher = "Royal Society Publishing",
number = "1",

}

TY - JOUR

T1 - Reverse engineering gene networks using global-local shrinkage rules

AU - Panchal, Viral

AU - Linder, Daniel F.

PY - 2020/2/6

Y1 - 2020/2/6

N2 - Inferring gene regulatory networks from high-throughput 'omics' data has proven to be a computationally demanding task of critical importance. Frequently, the classical methods break down owing to the curse of dimensionality, and popular strategies to overcome this are typically based on regularized versions of the classical methods. However, these approaches rely on loss functions that may not be robust and usually do not allow for the incorporation of prior information in a straightforward way. Fully Bayesian methods are equipped to handle both of these shortcomings quite naturally, and they offer the potential for improvements in network structure learning. We propose a Bayesian hierarchical model to reconstruct gene regulatory networks from time-series gene expression data, such as those common in perturbation experiments of biological systems. The proposed methodology uses global-local shrinkage priors for posterior selection of regulatory edges and relaxes the common normal likelihood assumption in order to allow for heavy-tailed data, which were shown in several of the cited references to severely impact network inference. We provide a sufficient condition for posterior propriety and derive an efficient Markov chain Monte Carlo via Gibbs sampling in the electronic supplementary material. We describe a novel way to detect multiple scales based on the corresponding posterior quantities. Finally, we demonstrate the performance of our approach in a simulation study and compare it with existing methods on real data from a T-cell activation study.

AB - Inferring gene regulatory networks from high-throughput 'omics' data has proven to be a computationally demanding task of critical importance. Frequently, the classical methods break down owing to the curse of dimensionality, and popular strategies to overcome this are typically based on regularized versions of the classical methods. However, these approaches rely on loss functions that may not be robust and usually do not allow for the incorporation of prior information in a straightforward way. Fully Bayesian methods are equipped to handle both of these shortcomings quite naturally, and they offer the potential for improvements in network structure learning. We propose a Bayesian hierarchical model to reconstruct gene regulatory networks from time-series gene expression data, such as those common in perturbation experiments of biological systems. The proposed methodology uses global-local shrinkage priors for posterior selection of regulatory edges and relaxes the common normal likelihood assumption in order to allow for heavy-tailed data, which were shown in several of the cited references to severely impact network inference. We provide a sufficient condition for posterior propriety and derive an efficient Markov chain Monte Carlo via Gibbs sampling in the electronic supplementary material. We describe a novel way to detect multiple scales based on the corresponding posterior quantities. Finally, we demonstrate the performance of our approach in a simulation study and compare it with existing methods on real data from a T-cell activation study.

KW - Bayesian shrinkage

KW - Gene networks

KW - Horseshoe prior

KW - Reverse engineering

UR - http://www.scopus.com/inward/record.url?scp=85077437709&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85077437709&partnerID=8YFLogxK

U2 - 10.1098/rsfs.2019.0049

DO - 10.1098/rsfs.2019.0049

M3 - Article

AN - SCOPUS:85077437709

VL - 10

JO - Interface Focus

JF - Interface Focus

SN - 2042-8898

IS - 1

M1 - 20190049

ER -