Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments

Stefano Parodi, Vito Pistoia, Marco Muselli

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

Abstract

TNRC (Test for Not Proper ROC Curve) is a statistical tool recently developed to identify differently expressed genes in microarray studies. In previous investigations it was demonstrated to be able to separate hidden subgroups in a two-class experiment, but being a univariate technique it could not exploit the complex multivariate correlation naturally occurring in gene expression data. In this study we show as the combination of TNRC with a standard technique of hierarchical clustering may provide useful biological insights. An example is provided using data from a publicly available data set of 4026 gene expression profiles in 42 samples of lymphomas and 14 samples of normal B cells.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages238-247
Number of pages10
Volume8452 LNBI
ISBN (Print)9783319090412
DOIs
Publication statusPublished - 2014
Event10th International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics, CIBB 2013 - Nice, France
Duration: Jun 20 2013Jun 22 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8452 LNBI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other10th International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics, CIBB 2013
CountryFrance
CityNice
Period6/20/136/22/13

Fingerprint

Receiver Operating Characteristic Curve
Hierarchical Clustering
Microarrays
Gene expression
Microarray
Genes
Gene
Gene Expression Profile
B Cells
Gene Expression Data
Univariate
Experiment
Experiments
Cells
Subgroup
Class
Standards

Keywords

  • Feature selection
  • Gene expression
  • Hierarchical clustering
  • ROC analysis

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Parodi, S., Pistoia, V., & Muselli, M. (2014). Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8452 LNBI, pp. 238-247). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8452 LNBI). Springer Verlag. https://doi.org/10.1007/978-3-319-09042-9_17

Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments. / Parodi, Stefano; Pistoia, Vito; Muselli, Marco.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8452 LNBI Springer Verlag, 2014. p. 238-247 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8452 LNBI).

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

Parodi, S, Pistoia, V & Muselli, M 2014, Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 8452 LNBI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8452 LNBI, Springer Verlag, pp. 238-247, 10th International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics, CIBB 2013, Nice, France, 6/20/13. https://doi.org/10.1007/978-3-319-09042-9_17
Parodi S, Pistoia V, Muselli M. Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8452 LNBI. Springer Verlag. 2014. p. 238-247. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-09042-9_17
Parodi, Stefano ; Pistoia, Vito ; Muselli, Marco. / Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8452 LNBI Springer Verlag, 2014. pp. 238-247 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{472629e827eb435f9b91c0fe305c401c,
title = "Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments",
abstract = "TNRC (Test for Not Proper ROC Curve) is a statistical tool recently developed to identify differently expressed genes in microarray studies. In previous investigations it was demonstrated to be able to separate hidden subgroups in a two-class experiment, but being a univariate technique it could not exploit the complex multivariate correlation naturally occurring in gene expression data. In this study we show as the combination of TNRC with a standard technique of hierarchical clustering may provide useful biological insights. An example is provided using data from a publicly available data set of 4026 gene expression profiles in 42 samples of lymphomas and 14 samples of normal B cells.",
keywords = "Feature selection, Gene expression, Hierarchical clustering, ROC analysis",
author = "Stefano Parodi and Vito Pistoia and Marco Muselli",
year = "2014",
doi = "10.1007/978-3-319-09042-9_17",
language = "English",
isbn = "9783319090412",
volume = "8452 LNBI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "238--247",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Combining not-proper ROC curves and hierarchical clustering to detect differentially expressed genes in microarray experiments

AU - Parodi, Stefano

AU - Pistoia, Vito

AU - Muselli, Marco

PY - 2014

Y1 - 2014

N2 - TNRC (Test for Not Proper ROC Curve) is a statistical tool recently developed to identify differently expressed genes in microarray studies. In previous investigations it was demonstrated to be able to separate hidden subgroups in a two-class experiment, but being a univariate technique it could not exploit the complex multivariate correlation naturally occurring in gene expression data. In this study we show as the combination of TNRC with a standard technique of hierarchical clustering may provide useful biological insights. An example is provided using data from a publicly available data set of 4026 gene expression profiles in 42 samples of lymphomas and 14 samples of normal B cells.

AB - TNRC (Test for Not Proper ROC Curve) is a statistical tool recently developed to identify differently expressed genes in microarray studies. In previous investigations it was demonstrated to be able to separate hidden subgroups in a two-class experiment, but being a univariate technique it could not exploit the complex multivariate correlation naturally occurring in gene expression data. In this study we show as the combination of TNRC with a standard technique of hierarchical clustering may provide useful biological insights. An example is provided using data from a publicly available data set of 4026 gene expression profiles in 42 samples of lymphomas and 14 samples of normal B cells.

KW - Feature selection

KW - Gene expression

KW - Hierarchical clustering

KW - ROC analysis

UR - http://www.scopus.com/inward/record.url?scp=84958538819&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958538819&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-09042-9_17

DO - 10.1007/978-3-319-09042-9_17

M3 - Conference contribution

AN - SCOPUS:84958538819

SN - 9783319090412

VL - 8452 LNBI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 238

EP - 247

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -