Selection of negative examples in learning gene regulatory networks

Michele Ceccarelli, Luigi Cerulo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

Supervised learning methods have been recently exploited to learn gene regulatory networks from gene expression data. They consist of building a binary classifier from feature vectors composed by expression levels of a set of known regulatory connections, available in public databases (eg. RegulonDB, TRRD, Transfac, IPA), and using such a classifier to predict new unknown connections. The input to a binary supervised classifier consists normally of positive and negative examples, but usually the only available information are a partial set of gene regulations, i.e. positive examples, and unlabeled data which could include both positive and negative examples. A fundamental challenge is the choice of negative examples from such unlabeled data to make the classifier able to learn from data. We exploit the known topology of a gene network to select such negative examples and show whether such an assumption benefits the performance of a classifier.

Original languageEnglish
Title of host publicationProceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009
Pages56-61
Number of pages6
DOIs
Publication statusPublished - 1 Dec 2009
Externally publishedYes
Event2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009 - Washington, DC, United States
Duration: 1 Nov 20094 Nov 2009

Other

Other2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009
CountryUnited States
CityWashington, DC
Period1/11/094/11/09

Fingerprint

Gene Regulatory Networks
Classifiers
Genes
Learning
Regulator Genes
Gene expression
Databases
Gene Expression
Supervised learning
Topology

Keywords

  • Gene regulatory networks
  • Machine learning

ASJC Scopus subject areas

  • Biomedical Engineering
  • Health Informatics
  • Health Information Management

Cite this

Ceccarelli, M., & Cerulo, L. (2009). Selection of negative examples in learning gene regulatory networks. In Proceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009 (pp. 56-61). [5332137] https://doi.org/10.1109/BIBMW.2009.5332137

Selection of negative examples in learning gene regulatory networks. / Ceccarelli, Michele; Cerulo, Luigi.

Proceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009. 2009. p. 56-61 5332137.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ceccarelli, M & Cerulo, L 2009, Selection of negative examples in learning gene regulatory networks. in Proceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009., 5332137, pp. 56-61, 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009, Washington, DC, United States, 1/11/09. https://doi.org/10.1109/BIBMW.2009.5332137
Ceccarelli M, Cerulo L. Selection of negative examples in learning gene regulatory networks. In Proceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009. 2009. p. 56-61. 5332137 https://doi.org/10.1109/BIBMW.2009.5332137
Ceccarelli, Michele ; Cerulo, Luigi. / Selection of negative examples in learning gene regulatory networks. Proceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009. 2009. pp. 56-61
@inproceedings{6fe2171dbf7a4fcea3f14e53375a7c67,
title = "Selection of negative examples in learning gene regulatory networks",
abstract = "Supervised learning methods have been recently exploited to learn gene regulatory networks from gene expression data. They consist of building a binary classifier from feature vectors composed by expression levels of a set of known regulatory connections, available in public databases (eg. RegulonDB, TRRD, Transfac, IPA), and using such a classifier to predict new unknown connections. The input to a binary supervised classifier consists normally of positive and negative examples, but usually the only available information are a partial set of gene regulations, i.e. positive examples, and unlabeled data which could include both positive and negative examples. A fundamental challenge is the choice of negative examples from such unlabeled data to make the classifier able to learn from data. We exploit the known topology of a gene network to select such negative examples and show whether such an assumption benefits the performance of a classifier.",
keywords = "Gene regulatory networks, Machine learning",
author = "Michele Ceccarelli and Luigi Cerulo",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/BIBMW.2009.5332137",
language = "English",
isbn = "9781424451210",
pages = "56--61",
booktitle = "Proceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009",

}

TY - GEN

T1 - Selection of negative examples in learning gene regulatory networks

AU - Ceccarelli, Michele

AU - Cerulo, Luigi

PY - 2009/12/1

Y1 - 2009/12/1

N2 - Supervised learning methods have been recently exploited to learn gene regulatory networks from gene expression data. They consist of building a binary classifier from feature vectors composed by expression levels of a set of known regulatory connections, available in public databases (eg. RegulonDB, TRRD, Transfac, IPA), and using such a classifier to predict new unknown connections. The input to a binary supervised classifier consists normally of positive and negative examples, but usually the only available information are a partial set of gene regulations, i.e. positive examples, and unlabeled data which could include both positive and negative examples. A fundamental challenge is the choice of negative examples from such unlabeled data to make the classifier able to learn from data. We exploit the known topology of a gene network to select such negative examples and show whether such an assumption benefits the performance of a classifier.

AB - Supervised learning methods have been recently exploited to learn gene regulatory networks from gene expression data. They consist of building a binary classifier from feature vectors composed by expression levels of a set of known regulatory connections, available in public databases (eg. RegulonDB, TRRD, Transfac, IPA), and using such a classifier to predict new unknown connections. The input to a binary supervised classifier consists normally of positive and negative examples, but usually the only available information are a partial set of gene regulations, i.e. positive examples, and unlabeled data which could include both positive and negative examples. A fundamental challenge is the choice of negative examples from such unlabeled data to make the classifier able to learn from data. We exploit the known topology of a gene network to select such negative examples and show whether such an assumption benefits the performance of a classifier.

KW - Gene regulatory networks

KW - Machine learning

UR - http://www.scopus.com/inward/record.url?scp=72949116439&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=72949116439&partnerID=8YFLogxK

U2 - 10.1109/BIBMW.2009.5332137

DO - 10.1109/BIBMW.2009.5332137

M3 - Conference contribution

SN - 9781424451210

SP - 56

EP - 61

BT - Proceedings - 2009 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2009

ER -