Integrating multiple-platform expression data through gene set features

Matej Holec, Filip Zelezny, Jiri Klema, Jakub Tolar

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

7 Citations (Scopus)

Abstract

We demonstrate a set-level approach to the integration of multiple platform gene expression data for predictive classification and show its utility for boosting classification performance when singleplatform samples are rare. We explore three ways of defining gene sets, including a novel way based on the notion of a fully coupled .flux related to metabolic pathways. In two tissue classification tasks, we empirically show that the gene set based approach is useful for combining heterogeneous expression data, while surprisingly, in experiments constrained to a single platform, biologically meaningful gene sets acting as sample features are often outperformed by random gene sets with no biological relevance.

Original languageEnglish (US)
Title of host publicationBioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings
Pages5-17
Number of pages13
DOIs
StatePublished - Jul 15 2009
Event5th International Symposium on Bioinformatics Research and Applications, ISBRA 2009 - Fort Lauderdale, FL, United States
Duration: May 13 2009May 16 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5542
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other5th International Symposium on Bioinformatics Research and Applications, ISBRA 2009
CountryUnited States
CityFort Lauderdale, FL
Period5/13/095/16/09

Fingerprint

Genes
Gene
Level-set Approach
Boosting
Gene Expression Data
Gene expression
Pathway
Tissue
Fluxes
Demonstrate
Experiment
Experiments

Cite this

Holec, M., Zelezny, F., Klema, J., & Tolar, J. (2009). Integrating multiple-platform expression data through gene set features. In Bioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings (pp. 5-17). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5542). https://doi.org/10.1007/978-3-642-01551-9_2

Integrating multiple-platform expression data through gene set features. / Holec, Matej; Zelezny, Filip; Klema, Jiri; Tolar, Jakub.

Bioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings. 2009. p. 5-17 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5542).

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

Holec, M, Zelezny, F, Klema, J & Tolar, J 2009, Integrating multiple-platform expression data through gene set features. in Bioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5542, pp. 5-17, 5th International Symposium on Bioinformatics Research and Applications, ISBRA 2009, Fort Lauderdale, FL, United States, 5/13/09. https://doi.org/10.1007/978-3-642-01551-9_2
Holec M, Zelezny F, Klema J, Tolar J. Integrating multiple-platform expression data through gene set features. In Bioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings. 2009. p. 5-17. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-01551-9_2
Holec, Matej ; Zelezny, Filip ; Klema, Jiri ; Tolar, Jakub. / Integrating multiple-platform expression data through gene set features. Bioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings. 2009. pp. 5-17 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{6a6d1fab145044b7b91f92e77f463fee,
title = "Integrating multiple-platform expression data through gene set features",
abstract = "We demonstrate a set-level approach to the integration of multiple platform gene expression data for predictive classification and show its utility for boosting classification performance when singleplatform samples are rare. We explore three ways of defining gene sets, including a novel way based on the notion of a fully coupled .flux related to metabolic pathways. In two tissue classification tasks, we empirically show that the gene set based approach is useful for combining heterogeneous expression data, while surprisingly, in experiments constrained to a single platform, biologically meaningful gene sets acting as sample features are often outperformed by random gene sets with no biological relevance.",
author = "Matej Holec and Filip Zelezny and Jiri Klema and Jakub Tolar",
year = "2009",
month = "7",
day = "15",
doi = "10.1007/978-3-642-01551-9_2",
language = "English (US)",
isbn = "9783642015502",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "5--17",
booktitle = "Bioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings",

}

TY - GEN

T1 - Integrating multiple-platform expression data through gene set features

AU - Holec, Matej

AU - Zelezny, Filip

AU - Klema, Jiri

AU - Tolar, Jakub

PY - 2009/7/15

Y1 - 2009/7/15

N2 - We demonstrate a set-level approach to the integration of multiple platform gene expression data for predictive classification and show its utility for boosting classification performance when singleplatform samples are rare. We explore three ways of defining gene sets, including a novel way based on the notion of a fully coupled .flux related to metabolic pathways. In two tissue classification tasks, we empirically show that the gene set based approach is useful for combining heterogeneous expression data, while surprisingly, in experiments constrained to a single platform, biologically meaningful gene sets acting as sample features are often outperformed by random gene sets with no biological relevance.

AB - We demonstrate a set-level approach to the integration of multiple platform gene expression data for predictive classification and show its utility for boosting classification performance when singleplatform samples are rare. We explore three ways of defining gene sets, including a novel way based on the notion of a fully coupled .flux related to metabolic pathways. In two tissue classification tasks, we empirically show that the gene set based approach is useful for combining heterogeneous expression data, while surprisingly, in experiments constrained to a single platform, biologically meaningful gene sets acting as sample features are often outperformed by random gene sets with no biological relevance.

UR - http://www.scopus.com/inward/record.url?scp=67650120989&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67650120989&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-01551-9_2

DO - 10.1007/978-3-642-01551-9_2

M3 - Conference contribution

SN - 9783642015502

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

SP - 5

EP - 17

BT - Bioinformatics Research and Applications - 5th International Symposium, ISBRA 2009, Proceedings

ER -