Biomarker identification by knowledge-driven multi-scale independent component analysis

Li Chen, Jianhua Xuan, Robert Clarke, Yue Wang

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

Abstract

Many statistical methods have been proposed to identify biomarkers from gene expression profiles. However, from expression data alone, statistical methods often fail to identify biologically meaningful biomarkers related to a specific biological process or disease under study. In this paper, we develop a novel strategy, namely knowledge-driven multi-scale independent component analysis (ICA), to infer regulatory signals and identify biologically relevant biomarkers from microarray data. Specifically, based on partial prior knowledge and clustering results, we apply ICA to find the most knowledge relevant linear regulatory mode in each subset of genes and then extract associated biomarkers according to their weighted loading factors. We have applied our method to a yeast cell cycle microarray dataset to find cell cycle regulated biomarkers. The experimental results indicate that our knowledge-driven multi-scale ICA method outperforms both baseline ICA method and correlation method significantly.

Original languageEnglish (US)
Title of host publication2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA
Pages261-264
Number of pages4
DOIs
StatePublished - 2008
Externally publishedYes
Event2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA - Bethesda, MD, United States
Duration: Nov 8 2007Nov 9 2007

Other

Other2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA
CountryUnited States
CityBethesda, MD
Period11/8/0711/9/07

Fingerprint

Independent component analysis
Biomarkers
Microarrays
Statistical methods
Cells
Correlation methods
Gene expression
Yeast
Genes

ASJC Scopus subject areas

  • Computer Science Applications
  • Information Systems

Cite this

Chen, L., Xuan, J., Clarke, R., & Wang, Y. (2008). Biomarker identification by knowledge-driven multi-scale independent component analysis. In 2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA (pp. 261-264). [4400934] https://doi.org/10.1109/LSSA.2007.4400934

Biomarker identification by knowledge-driven multi-scale independent component analysis. / Chen, Li; Xuan, Jianhua; Clarke, Robert; Wang, Yue.

2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA. 2008. p. 261-264 4400934.

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

Chen, L, Xuan, J, Clarke, R & Wang, Y 2008, Biomarker identification by knowledge-driven multi-scale independent component analysis. in 2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA., 4400934, pp. 261-264, 2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA, Bethesda, MD, United States, 11/8/07. https://doi.org/10.1109/LSSA.2007.4400934
Chen L, Xuan J, Clarke R, Wang Y. Biomarker identification by knowledge-driven multi-scale independent component analysis. In 2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA. 2008. p. 261-264. 4400934 https://doi.org/10.1109/LSSA.2007.4400934
Chen, Li ; Xuan, Jianhua ; Clarke, Robert ; Wang, Yue. / Biomarker identification by knowledge-driven multi-scale independent component analysis. 2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA. 2008. pp. 261-264
@inproceedings{9f4ded9001c54700b4c0bdf9ed3432b6,
title = "Biomarker identification by knowledge-driven multi-scale independent component analysis",
abstract = "Many statistical methods have been proposed to identify biomarkers from gene expression profiles. However, from expression data alone, statistical methods often fail to identify biologically meaningful biomarkers related to a specific biological process or disease under study. In this paper, we develop a novel strategy, namely knowledge-driven multi-scale independent component analysis (ICA), to infer regulatory signals and identify biologically relevant biomarkers from microarray data. Specifically, based on partial prior knowledge and clustering results, we apply ICA to find the most knowledge relevant linear regulatory mode in each subset of genes and then extract associated biomarkers according to their weighted loading factors. We have applied our method to a yeast cell cycle microarray dataset to find cell cycle regulated biomarkers. The experimental results indicate that our knowledge-driven multi-scale ICA method outperforms both baseline ICA method and correlation method significantly.",
author = "Li Chen and Jianhua Xuan and Robert Clarke and Yue Wang",
year = "2008",
doi = "10.1109/LSSA.2007.4400934",
language = "English (US)",
isbn = "9781424418138",
pages = "261--264",
booktitle = "2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA",

}

TY - GEN

T1 - Biomarker identification by knowledge-driven multi-scale independent component analysis

AU - Chen, Li

AU - Xuan, Jianhua

AU - Clarke, Robert

AU - Wang, Yue

PY - 2008

Y1 - 2008

N2 - Many statistical methods have been proposed to identify biomarkers from gene expression profiles. However, from expression data alone, statistical methods often fail to identify biologically meaningful biomarkers related to a specific biological process or disease under study. In this paper, we develop a novel strategy, namely knowledge-driven multi-scale independent component analysis (ICA), to infer regulatory signals and identify biologically relevant biomarkers from microarray data. Specifically, based on partial prior knowledge and clustering results, we apply ICA to find the most knowledge relevant linear regulatory mode in each subset of genes and then extract associated biomarkers according to their weighted loading factors. We have applied our method to a yeast cell cycle microarray dataset to find cell cycle regulated biomarkers. The experimental results indicate that our knowledge-driven multi-scale ICA method outperforms both baseline ICA method and correlation method significantly.

AB - Many statistical methods have been proposed to identify biomarkers from gene expression profiles. However, from expression data alone, statistical methods often fail to identify biologically meaningful biomarkers related to a specific biological process or disease under study. In this paper, we develop a novel strategy, namely knowledge-driven multi-scale independent component analysis (ICA), to infer regulatory signals and identify biologically relevant biomarkers from microarray data. Specifically, based on partial prior knowledge and clustering results, we apply ICA to find the most knowledge relevant linear regulatory mode in each subset of genes and then extract associated biomarkers according to their weighted loading factors. We have applied our method to a yeast cell cycle microarray dataset to find cell cycle regulated biomarkers. The experimental results indicate that our knowledge-driven multi-scale ICA method outperforms both baseline ICA method and correlation method significantly.

UR - http://www.scopus.com/inward/record.url?scp=50949099313&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50949099313&partnerID=8YFLogxK

U2 - 10.1109/LSSA.2007.4400934

DO - 10.1109/LSSA.2007.4400934

M3 - Conference contribution

AN - SCOPUS:50949099313

SN - 9781424418138

SP - 261

EP - 264

BT - 2007 IEEE/NIH Life Science Systems and Applications Workshop, LISA

ER -