Robust data-driven incorporation of prior knowledge into the inference of dynamic regulatory networks

Alex Greenfield, Christoph Hafemeister, Richard Bonneau

Research output: Contribution to journalArticle

Abstract

Inferring global regulatory networks (GRNs) from genome-wide data is a computational challenge central to the field of systems biology. Although the primary data currently used to infer GRNs consist of gene expression and proteomics measurements, there is a growing abundance of alternate data types that can reveal regulatory interactions, e.g. ChIP-Chip, literature-derived interactions, protein-protein interactions. GRN inference requires the development of integrative methods capable of using these alternate data as priors on the GRN structure. Each source of structure priors has its unique biases and inherent potential errors; thus, GRN methods using these data must be robust to noisy inputs. Results:We developed two methods for incorporating structure priors into GRN inference. Both methods [Modified Elastic Net (MEN) and Bayesian Best Subset Regression (BBSR)] extend the previously described Inferelator framework, enabling the use of prior information. We test our methods on one synthetic and two bacterial datasets, and show that both MEN and BBSR infer accurate GRNs even when the structure prior used has significant amounts of error (490%erroneous interactions). We find that BBSR outperforms MEN at inferring GRNs from expression data and noisy structure priors.

Original languageEnglish (US)
Pages (from-to)1060-1067
Number of pages8
JournalBioinformatics
Volume29
Issue number8
DOIs
StatePublished - Apr 15 2013

Fingerprint

Regulatory Networks
Dynamic Networks
Prior Knowledge
Data-driven
Proteins
Elastic Net
Gene expression
Genes
Regression
Systems Biology
Alternate
Subset
Chip
Proteomics
Interaction
Genome
Protein-protein Interaction
Prior Information
Gene Expression
Network Structure

ASJC Scopus subject areas

  • Biochemistry
  • Molecular Biology
  • Computational Theory and Mathematics
  • Computer Science Applications
  • Computational Mathematics
  • Statistics and Probability
  • Medicine(all)

Cite this

Robust data-driven incorporation of prior knowledge into the inference of dynamic regulatory networks. / Greenfield, Alex; Hafemeister, Christoph; Bonneau, Richard.

In: Bioinformatics, Vol. 29, No. 8, 15.04.2013, p. 1060-1067.

Research output: Contribution to journalArticle

Greenfield, Alex ; Hafemeister, Christoph ; Bonneau, Richard. / Robust data-driven incorporation of prior knowledge into the inference of dynamic regulatory networks. In: Bioinformatics. 2013 ; Vol. 29, No. 8. pp. 1060-1067.
@article{35bd554ba1ba40988b55eb480b125baa,
title = "Robust data-driven incorporation of prior knowledge into the inference of dynamic regulatory networks",
abstract = "Inferring global regulatory networks (GRNs) from genome-wide data is a computational challenge central to the field of systems biology. Although the primary data currently used to infer GRNs consist of gene expression and proteomics measurements, there is a growing abundance of alternate data types that can reveal regulatory interactions, e.g. ChIP-Chip, literature-derived interactions, protein-protein interactions. GRN inference requires the development of integrative methods capable of using these alternate data as priors on the GRN structure. Each source of structure priors has its unique biases and inherent potential errors; thus, GRN methods using these data must be robust to noisy inputs. Results:We developed two methods for incorporating structure priors into GRN inference. Both methods [Modified Elastic Net (MEN) and Bayesian Best Subset Regression (BBSR)] extend the previously described Inferelator framework, enabling the use of prior information. We test our methods on one synthetic and two bacterial datasets, and show that both MEN and BBSR infer accurate GRNs even when the structure prior used has significant amounts of error (490{\%}erroneous interactions). We find that BBSR outperforms MEN at inferring GRNs from expression data and noisy structure priors.",
author = "Alex Greenfield and Christoph Hafemeister and Richard Bonneau",
year = "2013",
month = "4",
day = "15",
doi = "10.1093/bioinformatics/btt099",
language = "English (US)",
volume = "29",
pages = "1060--1067",
journal = "Bioinformatics",
issn = "1367-4803",
publisher = "Oxford University Press",
number = "8",

}

TY - JOUR

T1 - Robust data-driven incorporation of prior knowledge into the inference of dynamic regulatory networks

AU - Greenfield, Alex

AU - Hafemeister, Christoph

AU - Bonneau, Richard

PY - 2013/4/15

Y1 - 2013/4/15

N2 - Inferring global regulatory networks (GRNs) from genome-wide data is a computational challenge central to the field of systems biology. Although the primary data currently used to infer GRNs consist of gene expression and proteomics measurements, there is a growing abundance of alternate data types that can reveal regulatory interactions, e.g. ChIP-Chip, literature-derived interactions, protein-protein interactions. GRN inference requires the development of integrative methods capable of using these alternate data as priors on the GRN structure. Each source of structure priors has its unique biases and inherent potential errors; thus, GRN methods using these data must be robust to noisy inputs. Results:We developed two methods for incorporating structure priors into GRN inference. Both methods [Modified Elastic Net (MEN) and Bayesian Best Subset Regression (BBSR)] extend the previously described Inferelator framework, enabling the use of prior information. We test our methods on one synthetic and two bacterial datasets, and show that both MEN and BBSR infer accurate GRNs even when the structure prior used has significant amounts of error (490%erroneous interactions). We find that BBSR outperforms MEN at inferring GRNs from expression data and noisy structure priors.

AB - Inferring global regulatory networks (GRNs) from genome-wide data is a computational challenge central to the field of systems biology. Although the primary data currently used to infer GRNs consist of gene expression and proteomics measurements, there is a growing abundance of alternate data types that can reveal regulatory interactions, e.g. ChIP-Chip, literature-derived interactions, protein-protein interactions. GRN inference requires the development of integrative methods capable of using these alternate data as priors on the GRN structure. Each source of structure priors has its unique biases and inherent potential errors; thus, GRN methods using these data must be robust to noisy inputs. Results:We developed two methods for incorporating structure priors into GRN inference. Both methods [Modified Elastic Net (MEN) and Bayesian Best Subset Regression (BBSR)] extend the previously described Inferelator framework, enabling the use of prior information. We test our methods on one synthetic and two bacterial datasets, and show that both MEN and BBSR infer accurate GRNs even when the structure prior used has significant amounts of error (490%erroneous interactions). We find that BBSR outperforms MEN at inferring GRNs from expression data and noisy structure priors.

UR - http://www.scopus.com/inward/record.url?scp=84876207916&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84876207916&partnerID=8YFLogxK

U2 - 10.1093/bioinformatics/btt099

DO - 10.1093/bioinformatics/btt099

M3 - Article

VL - 29

SP - 1060

EP - 1067

JO - Bioinformatics

JF - Bioinformatics

SN - 1367-4803

IS - 8

ER -