Selecting genes by test statistics

Dechang Chen, Zhenqiu Liu, Xiaobin Ma, Dong Hua

Research output: Contribution to journalArticle

44 Citations (Scopus)

Abstract

Gene selection is an important issue in analyzing multiclass microarray data. Among many proposed selection methods, the traditional ANOVA F test statistic has been employed to identify informative genes for both class prediction (classification) and discovery problems. However, the F test statistic assumes an equal variance. This assumption may not be realistic for gene expression data. This paper explores other alternative test statistics which can handle heterogeneity of the variances. We study five such test statistics, which include Brown-Forsythe test statistic and Welch test statistic. Their performance is evaluated and compared with that of F statistic over different classification methods applied to publicly available microarray datasets.

Original languageEnglish (US)
Pages (from-to)132-138
Number of pages7
JournalJournal of Biomedicine and Biotechnology
Volume2005
Issue number2
DOIs
StatePublished - Jun 30 2005

Fingerprint

Genes
Statistics
Microarrays
Analysis of variance (ANOVA)
Gene expression
Analysis of Variance
Gene Expression

All Science Journal Classification (ASJC) codes

  • Biotechnology
  • Molecular Medicine
  • Molecular Biology
  • Genetics
  • Health, Toxicology and Mutagenesis

Cite this

Chen, Dechang ; Liu, Zhenqiu ; Ma, Xiaobin ; Hua, Dong. / Selecting genes by test statistics. In: Journal of Biomedicine and Biotechnology. 2005 ; Vol. 2005, No. 2. pp. 132-138.
@article{54e0848ed31747739bfbb3a9b1d200ba,
title = "Selecting genes by test statistics",
abstract = "Gene selection is an important issue in analyzing multiclass microarray data. Among many proposed selection methods, the traditional ANOVA F test statistic has been employed to identify informative genes for both class prediction (classification) and discovery problems. However, the F test statistic assumes an equal variance. This assumption may not be realistic for gene expression data. This paper explores other alternative test statistics which can handle heterogeneity of the variances. We study five such test statistics, which include Brown-Forsythe test statistic and Welch test statistic. Their performance is evaluated and compared with that of F statistic over different classification methods applied to publicly available microarray datasets.",
author = "Dechang Chen and Zhenqiu Liu and Xiaobin Ma and Dong Hua",
year = "2005",
month = "6",
day = "30",
doi = "10.1155/JBB.2005.132",
language = "English (US)",
volume = "2005",
pages = "132--138",
journal = "BioMed Research International",
issn = "2314-6133",
publisher = "Hindawi Publishing Corporation",
number = "2",

}

Selecting genes by test statistics. / Chen, Dechang; Liu, Zhenqiu; Ma, Xiaobin; Hua, Dong.

In: Journal of Biomedicine and Biotechnology, Vol. 2005, No. 2, 30.06.2005, p. 132-138.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Selecting genes by test statistics

AU - Chen, Dechang

AU - Liu, Zhenqiu

AU - Ma, Xiaobin

AU - Hua, Dong

PY - 2005/6/30

Y1 - 2005/6/30

N2 - Gene selection is an important issue in analyzing multiclass microarray data. Among many proposed selection methods, the traditional ANOVA F test statistic has been employed to identify informative genes for both class prediction (classification) and discovery problems. However, the F test statistic assumes an equal variance. This assumption may not be realistic for gene expression data. This paper explores other alternative test statistics which can handle heterogeneity of the variances. We study five such test statistics, which include Brown-Forsythe test statistic and Welch test statistic. Their performance is evaluated and compared with that of F statistic over different classification methods applied to publicly available microarray datasets.

AB - Gene selection is an important issue in analyzing multiclass microarray data. Among many proposed selection methods, the traditional ANOVA F test statistic has been employed to identify informative genes for both class prediction (classification) and discovery problems. However, the F test statistic assumes an equal variance. This assumption may not be realistic for gene expression data. This paper explores other alternative test statistics which can handle heterogeneity of the variances. We study five such test statistics, which include Brown-Forsythe test statistic and Welch test statistic. Their performance is evaluated and compared with that of F statistic over different classification methods applied to publicly available microarray datasets.

UR - http://www.scopus.com/inward/record.url?scp=27744557285&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=27744557285&partnerID=8YFLogxK

U2 - 10.1155/JBB.2005.132

DO - 10.1155/JBB.2005.132

M3 - Article

C2 - 16046818

AN - SCOPUS:27744557285

VL - 2005

SP - 132

EP - 138

JO - BioMed Research International

JF - BioMed Research International

SN - 2314-6133

IS - 2

ER -