Feature (Gene) Selection in Gene Expression-Based Tumor Classification

Xiong, Momiao; Li, Wuju; Zhao, Jinying; Li, Jin; Boerwinkle, Eric

doi:10.1006/mgme.2001.3193

Cited by 123 publications

(61 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A s explained by Xiong et al (1), there is increasing interest in changing the emphasis of tumor classification from morphologic to molecular. In this context, the problem is to construct a classifier or a prediction (discriminant) rule R that can accurately predict the class of origin of a tumor tissue with feature vector x, which is unclassified with respect to a known number g (Ն2) of distinct tissue types.…”

mentioning

confidence: 99%

Selection bias in gene extraction on the basis of microarray gene-expression data

Ambroise

McLachlan

2002

Proc. Natl. Acad. Sci. U.S.A.

1,234

923

View full text Add to dashboard Cite

In the context of cancer diagnosis and treatment, we consider the problem of constructing an accurate prediction rule on the basis of a relatively small number of tumor tissue samples of known type containing the expression data on very many (possibly thousands) genes. Recently, results have been presented in the literature suggesting that it is possible to construct a prediction rule from only a few genes such that it has a negligible prediction error rate. However, in these results the test error or the leave-one-out cross-validated error is calculated without allowance for the selection bias. There is no allowance because the rule is either tested on tissue samples that were used in the first instance to select the genes being used in the rule or because the cross-validation of the rule is not external to the selection process; that is, gene selection is not performed in training the rule at each stage of the crossvalidation process. We describe how in practice the selection bias can be assessed and corrected for by either performing a crossvalidation or applying the bootstrap external to the selection process. We recommend using 10-fold rather than leave-one-out cross-validation, and concerning the bootstrap, we suggest using the so-called .632؉ bootstrap error estimate designed to handle overfitted prediction rules. Using two published data sets, we demonstrate that when correction is made for the selection bias, the cross-validated error is no longer zero for a subset of only a few genes.

show abstract

mentioning

confidence: 99%

Selection bias in gene extraction on the basis of microarray gene-expression data

Ambroise

McLachlan

2002

Proc. Natl. Acad. Sci. U.S.A.

1,234

923

View full text Add to dashboard Cite

show abstract

“…In a two group setting the BW-ratio reduces to the same statistic as the t-test. The prediction strength (PS) (Xiong et al 2001) of a certain gene is defined as the ratio of the difference in mean log expression level between the two groups and the sum of the variances of the two classes. The between-class scatter score (BC-score) belongs to the class of correlation scores (Chai and Domenicioni 2004).…”

Section: Other Test Statisticsmentioning

confidence: 99%

Performance of Gene Selection and Classification Methods in a Microarray Setting: A Simulation Study

Sanden

Lin

Burzykowski

2008

Communications in Statistics - Simulation and Computation

View full text Add to dashboard Cite

In a previous paper we investigated the performance of several classification methods for cDNA-microarrays. Via simulations various experimental settings could be explored without having to conduct expensive microarray studies. For the selection of genes on which classification was based, one particular method was applied. Gene selection is however a very important aspect of classification. We would like to extend the previous study by considering several gene selection methods. Furthermore, the stability of the methods with respect to distributional assumptions is examined by also considering data simulated from a symmetric and asymmetric Laplace distribution in addition to normally distributed microarray data.

show abstract

“…Recent researches have shown that a small number of genes is sufficient for accurate diagnosis of most cancers, even though the number of genes vary greatly between different diseases [15]. Indeed, a large set of gene expression features will not only significantly bring higher computational cost and slow down the learning process, but also decrease the classification accuracy due to the phenomenon known as curse of dimensionality, in which the risk of over-fitting increases as the number of selected genes grows [15].…”

Section: Gene Selectionmentioning

confidence: 99%

“…Indeed, a large set of gene expression features will not only significantly bring higher computational cost and slow down the learning process, but also decrease the classification accuracy due to the phenomenon known as curse of dimensionality, in which the risk of over-fitting increases as the number of selected genes grows [15]. More importantly, by using a small subset of genes, we can not only get a better diagnostic accuracy, but also get an opportunity to further analyse the nature of the disease and the genetic mechanisms responsible for it.…”

Section: Gene Selectionmentioning

confidence: 99%

Neuro-Fuzzy Ensemble Approach for Microarray Cancer Gene Expression Data Analysis

Wang

Palade

Xu³

2006

2006 International Symposium on Evolving Fuzzy Systems

View full text Add to dashboard Cite

Abstract-A Neuro-Fuzzy Ensemble model (NFE) is proposed in this paper for analysing the gene expression data from microarray experiments. The proposed approach was tested on three benchmark cancer gene expression data sets. Experimental results show that our NFE model can be used as an efficient computational tool for microarray data analysis. In addition, compared to some current most widely used approaches, NeuroFuzzy(NF)-based models not only supply good classification results, but their behavior can also be explained and interpreted in human understandable terms, which provides the researchers with a better understanding of the data.

show abstract

Feature (Gene) Selection in Gene Expression-Based Tumor Classification

Cited by 123 publications

References 15 publications

Selection bias in gene extraction on the basis of microarray gene-expression data

Selection bias in gene extraction on the basis of microarray gene-expression data

Performance of Gene Selection and Classification Methods in a Microarray Setting: A Simulation Study

Neuro-Fuzzy Ensemble Approach for Microarray Cancer Gene Expression Data Analysis

Contact Info

Product

Resources

About