Abstract

Gene selection is an important task in bioinformatics studies, because the accuracy of cancer classification generally depends upon the genes that have biological relevance to the classifying problems. In this work, randomization test (RT) is used as a gene selection method for dealing with gene expression data. In the method, a statistic derived from the statistics of the regression coefficients in a series of partial least squares discriminant analysis (PLSDA) models is used to evaluate the significance of the genes. Informative genes are selected for classifying the four gene expression datasets of prostate cancer, lung cancer, leukemia and non-small cell lung cancer (NSCLC) and the rationality of the results is validated by multiple linear regression (MLR) modeling and principal component analysis (PCA). With the selected genes, satisfactory results can be obtained.

Introduction

Cancer classification based on microarray has become a popular research topic in bioinformatics, which can be used to detect subtypes of cancers and produce therapies. A great many of studies have appeared for cancer classification [1–۳]. These methods include principal component analysis (PCA) [4,5], k-nearest neighbor (k-NN) [6], hierarchical clustering analysis (HCA) [7], support vector machine (SVM) [8], Bayesian method [9], partial least squares discriminant analysis (PLSDA) [10], ensemble methods [11], etc. Among these methods, PLSDA has been the most commonly used one for cancer classification due to its simplicity [12–۱۴]. Moreover, as a dimension reduction technique, PLS has been used in gene expression data analysis even in the case where the number of genes exceeds the number of samples.