【論】Simon,2003,Pitfalls in the use of DNA micro～

Richard Simon, Michael D.Radmacher, Kevin Dobbin, Lisa M.McShane
Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification
Natl.Cancer Inst.95 (2003) 14.
[PDFダウンロード]

●目次
・Experimental objectives drive design and analysis
・Limitations of cluster analysis for class prediction
・Class prediction using supervised methods
・Cross-validation of prediction accuracy
・Validation on independent data
・Reporting the error rate
・Claims for new class prediction methods
・Conclusion

・注意点「Although cluster analysis appropriate for class discovery, it is often not effective for class comparison or class prediction.」
・問題点「Cluster analysis also does not provide statistically valid quantitative information about which genes are differentially expressed between classes.」
・問題点「One major limitation of supervised methods is overfitting the predictor.」
・結論「We recommend that supervised methods rather than cluster anlyses be used for class prediction and class comparison studies.」
・結論「If cross-validation is used to estimate prediction accuracy, then the entire model-building process, including the selection of informative genes, should be repeated in each cross-validation training set.」