Microarray-based cancer prediction using single genes

BMC Bioinformatics

Research Article

Richard Simon¹  Xiaosheng Wang¹ 

[1] Biometric Research Branch, National Cancer Institute, National Institutes of Health, 20852, Rockville, MD, USA;

关键词: Support Vector Machine;    Classification Accuracy;    Random Forest;    Cancer Dataset;    Random Forest Classifier;   

DOI  :  10.1186/1471-2105-12-391

 received in 2011-05-24, accepted in 2011-10-07,  发布年份 2011

来源: Springer

【 摘 要 】

BackgroundAlthough numerous methods of using microarray data analysis for cancer classification have been proposed, most utilize many genes to achieve accurate classification. This can hamper interpretability of the models and ease of translation to other assay platforms. We explored the use of single genes to construct classification models. We first identified the genes with the most powerful univariate class discrimination ability and then constructed simple classification rules for class prediction using the single genes.ResultsWe applied our model development algorithm to eleven cancer gene expression datasets and compared classification accuracy to that for standard methods including Diagonal Linear Discriminant Analysis, k-Nearest Neighbor, Support Vector Machine and Random Forest. The single gene classifiers provided classification accuracy comparable to or better than those obtained by existing methods in most cases. We analyzed the factors that determined when simple single gene classification is effective and when more complex modeling is warranted.ConclusionsFor most of the datasets examined, the single-gene classification methods appear to work as well as more standard methods, suggesting that simple models could perform well in microarray-based cancer prediction.

【 授权许可】

CC BY   
© Wang and Simon; licensee BioMed Central Ltd. 2011