Bioinformatics Advance Access published online on June 29, 2006
Bioinformatics, doi:10.1093/bioinformatics/btl339
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1 Department of Bioinformatics, Harbin Medical University, Harbin 150086, China
* To whom correspondence should be addressed.
Motivation: Microarrays datasets frequently contain a large number of missing values (MVs), which need to be estimated and replaced for subsequent data mining. The focus of the paper is to study the effects of different MV treatments for cDNA microarray data on disease classification analysis. Results: By analyzing five datasets, we demonstrate that among three kinds of classifiers evaluated in this study, support vector machine classifiers are robust to varied MV imputation methods (e.g. replacing MVs by zero, K nearest-neighbor imputation algorithm, local least square imputation, and Bayesian principal component analysis), while the classification and regression tree classifiers are sensitive in terms of classification accuracy. The K nearest-neighbor classifiers built on differentially expressed genes are robust to the varied MV treatments, but the performances of the K nearest-neighbor classifiers based on all measured genes can be significantly deteriorated when imputing MVs for genes with larger missing rate (MR) (e.g. MR>5%). Generally, while replacing MVs by zero performs relatively poor, the other imputation algorithms have little difference in affecting classification performances of the SVM or KNN classifiers. We further demonstrate the power and feasibility of our recently proposed functional expression profile approach as means to handle microarray data with MVs. The functional expression profiles, which are derived from the functional modules that are enriched with sets of differentially expressed genes and thus can be consistently identified under varied MV treatments, achieve precise disease classification with better biological interpretation. We conclude that the choice of MV treatments should be determined in context of the later approaches used for disease classification. The suggested exclusion criterion of ignoring the genes with larger MR (e.g. >5%), while justifiable for some classifiers such as KNN classifiers, might not be considered as a general rule for all classifiers.
Received May 25, 2006
Revised June 16, 2006
Accepted June 16, 2006
Article
Effects of replacing the unreliable cDNA microarray measurements on the disease classification based on gene expression profiles and functional modules
Dong Wang 1,
Yingli Lv 1,
Zheng Guo 2 *,
Xia Li 1,
Yanhui Li 1,
Jing Zhu 1,
Da Yang 1,
Jianzhen Xu 1,
Chenguang Wang 1,
Shaoqi Rao 3,
and
Baofeng Yang 4
2 Department of Bioinformatics, Harbin Medical University, Harbin 150086, China; Department of Pharmacology and Bio-pharmaceutical Key Laboratory of Heilongjiang Province and State, Harbin Medical University, Harbin 150086, China
3 Department of Bioinformatics, Harbin Medical University, Harbin 150086, China; Department of Molecular Cardiology and Department of Cardiovascular Medicine, the Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195, USA
4 Department of Pharmacology and Bio-pharmaceutical Key Laboratory of Heilongjiang Province and State, Harbin Medical University, Harbin 150086, China
Zheng Guo, E-mail: guoz{at}ems.hrbmu.edu.cn
Abstract
Associate Editor: Martin Bishop
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Yang, Y. Li, H. Xiao, Q. Liu, M. Zhang, J. Zhu, W. Ma, C. Yao, J. Wang, D. Wang, et al. Gaining confidence in biological interpretation of the microarray data: the functional consistence of the significant GO categories Bioinformatics, January 15, 2008; 24(2): 265 - 271. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Varshavsky, A. Gottlieb, D. Horn, and M. Linial Unsupervised feature selection under perturbations: meeting the challenges of biological data Bioinformatics, December 15, 2007; 23(24): 3343 - 3349. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Guo, Y. Li, X. Gong, C. Yao, W. Ma, D. Wang, Y. Li, J. Zhu, M. Zhang, D. Yang, et al. Edge-based scoring and searching method for identifying condition-responsive protein protein interaction sub-network Bioinformatics, August 15, 2007; 23(16): 2121 - 2128. [Abstract] [Full Text] [PDF]
|
|