Measuring similarities between gene expression profiles through new data transformations

Kyungpil Kim, Shibo Zhang, Keni Jiang, Li Cai, In Beum Lee, Lewis J. Feldman, Haiyan Huang

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

Background: Clustering methods are widely used on gene expression data to categorize genes with similar expression profiles. Finding an appropriate (dis)similarity measure is critical to the analysis. In our study, we developed a new measure for clustering the genes when the key factor is the shape of the profile, and when the expression magnitude should also be accounted for in determining the gene relationship. This is achieved by modeling the shape and magnitude parameters separately in a gene expression profile, and then using the estimated shape and magnitude parameters to define a measure in a new feature space. Results: We explored several different transformation schemes to construct the feature spaces that include a space whose features are determined by the mutual differences of the original expression components, a space derived from a parametric covariance matrix, and the principal component space in traditional PCA analysis. The former two are the newly proposed and the latter is explored for comparison purposes. The new measures we defined in these feature spaces were employed in a K-means clustering procedure to perform analyses. Applying these algorithms to a simulation dataset, a developing mouse retina SAGE dataset, a small yeast sporulation cDNA dataset, and a maize root affymetrix microarray dataset, we found from the results that the algorithm associated with the first feature space, named TransChisq, showed clear advantages over other methods. Conclusion: The proposed TransChisq is very promising in capturing meaningful gene expression clusters. This study also demonstrates the importance of data transformations in defining an efficient distance measure. Our method should provide new insights in analyzing gene expression data. The clustering algorithms are available upon request.

Original languageEnglish (US)
Article number29
JournalBMC bioinformatics
Volume8
DOIs
StatePublished - Mar 9 2007

Fingerprint

Data Transformation
Gene Expression Profile
Feature Space
Transcriptome
Gene expression
Cluster Analysis
Genes
Gene Expression
Gene Expression Data
Gene
Passive Cutaneous Anaphylaxis
Microarrays
Multigene Family
Covariance matrix
Clustering algorithms
Retina
Yeast
Maize
Zea mays
K-means Clustering

All Science Journal Classification (ASJC) codes

  • Applied Mathematics
  • Molecular Biology
  • Structural Biology
  • Biochemistry
  • Computer Science Applications

Cite this

Kim, Kyungpil ; Zhang, Shibo ; Jiang, Keni ; Cai, Li ; Lee, In Beum ; Feldman, Lewis J. ; Huang, Haiyan. / Measuring similarities between gene expression profiles through new data transformations. In: BMC bioinformatics. 2007 ; Vol. 8.
@article{7824e3fb1fae4fe982373cb3f813edf9,
title = "Measuring similarities between gene expression profiles through new data transformations",
abstract = "Background: Clustering methods are widely used on gene expression data to categorize genes with similar expression profiles. Finding an appropriate (dis)similarity measure is critical to the analysis. In our study, we developed a new measure for clustering the genes when the key factor is the shape of the profile, and when the expression magnitude should also be accounted for in determining the gene relationship. This is achieved by modeling the shape and magnitude parameters separately in a gene expression profile, and then using the estimated shape and magnitude parameters to define a measure in a new feature space. Results: We explored several different transformation schemes to construct the feature spaces that include a space whose features are determined by the mutual differences of the original expression components, a space derived from a parametric covariance matrix, and the principal component space in traditional PCA analysis. The former two are the newly proposed and the latter is explored for comparison purposes. The new measures we defined in these feature spaces were employed in a K-means clustering procedure to perform analyses. Applying these algorithms to a simulation dataset, a developing mouse retina SAGE dataset, a small yeast sporulation cDNA dataset, and a maize root affymetrix microarray dataset, we found from the results that the algorithm associated with the first feature space, named TransChisq, showed clear advantages over other methods. Conclusion: The proposed TransChisq is very promising in capturing meaningful gene expression clusters. This study also demonstrates the importance of data transformations in defining an efficient distance measure. Our method should provide new insights in analyzing gene expression data. The clustering algorithms are available upon request.",
author = "Kyungpil Kim and Shibo Zhang and Keni Jiang and Li Cai and Lee, {In Beum} and Feldman, {Lewis J.} and Haiyan Huang",
year = "2007",
month = "3",
day = "9",
doi = "https://doi.org/10.1186/1471-2105-8-29",
language = "English (US)",
volume = "8",
journal = "BMC Bioinformatics",
issn = "1471-2105",
publisher = "BioMed Central Ltd.",

}

Measuring similarities between gene expression profiles through new data transformations. / Kim, Kyungpil; Zhang, Shibo; Jiang, Keni; Cai, Li; Lee, In Beum; Feldman, Lewis J.; Huang, Haiyan.

In: BMC bioinformatics, Vol. 8, 29, 09.03.2007.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Measuring similarities between gene expression profiles through new data transformations

AU - Kim, Kyungpil

AU - Zhang, Shibo

AU - Jiang, Keni

AU - Cai, Li

AU - Lee, In Beum

AU - Feldman, Lewis J.

AU - Huang, Haiyan

PY - 2007/3/9

Y1 - 2007/3/9

N2 - Background: Clustering methods are widely used on gene expression data to categorize genes with similar expression profiles. Finding an appropriate (dis)similarity measure is critical to the analysis. In our study, we developed a new measure for clustering the genes when the key factor is the shape of the profile, and when the expression magnitude should also be accounted for in determining the gene relationship. This is achieved by modeling the shape and magnitude parameters separately in a gene expression profile, and then using the estimated shape and magnitude parameters to define a measure in a new feature space. Results: We explored several different transformation schemes to construct the feature spaces that include a space whose features are determined by the mutual differences of the original expression components, a space derived from a parametric covariance matrix, and the principal component space in traditional PCA analysis. The former two are the newly proposed and the latter is explored for comparison purposes. The new measures we defined in these feature spaces were employed in a K-means clustering procedure to perform analyses. Applying these algorithms to a simulation dataset, a developing mouse retina SAGE dataset, a small yeast sporulation cDNA dataset, and a maize root affymetrix microarray dataset, we found from the results that the algorithm associated with the first feature space, named TransChisq, showed clear advantages over other methods. Conclusion: The proposed TransChisq is very promising in capturing meaningful gene expression clusters. This study also demonstrates the importance of data transformations in defining an efficient distance measure. Our method should provide new insights in analyzing gene expression data. The clustering algorithms are available upon request.

AB - Background: Clustering methods are widely used on gene expression data to categorize genes with similar expression profiles. Finding an appropriate (dis)similarity measure is critical to the analysis. In our study, we developed a new measure for clustering the genes when the key factor is the shape of the profile, and when the expression magnitude should also be accounted for in determining the gene relationship. This is achieved by modeling the shape and magnitude parameters separately in a gene expression profile, and then using the estimated shape and magnitude parameters to define a measure in a new feature space. Results: We explored several different transformation schemes to construct the feature spaces that include a space whose features are determined by the mutual differences of the original expression components, a space derived from a parametric covariance matrix, and the principal component space in traditional PCA analysis. The former two are the newly proposed and the latter is explored for comparison purposes. The new measures we defined in these feature spaces were employed in a K-means clustering procedure to perform analyses. Applying these algorithms to a simulation dataset, a developing mouse retina SAGE dataset, a small yeast sporulation cDNA dataset, and a maize root affymetrix microarray dataset, we found from the results that the algorithm associated with the first feature space, named TransChisq, showed clear advantages over other methods. Conclusion: The proposed TransChisq is very promising in capturing meaningful gene expression clusters. This study also demonstrates the importance of data transformations in defining an efficient distance measure. Our method should provide new insights in analyzing gene expression data. The clustering algorithms are available upon request.

UR - http://www.scopus.com/inward/record.url?scp=33847657554&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33847657554&partnerID=8YFLogxK

U2 - https://doi.org/10.1186/1471-2105-8-29

DO - https://doi.org/10.1186/1471-2105-8-29

M3 - Article

VL - 8

JO - BMC Bioinformatics

JF - BMC Bioinformatics

SN - 1471-2105

M1 - 29

ER -