Detecting Outlier Samples in Microarray Data

摘要

In this paper, we address the problem of detecting outlier samples with highly different expres-nsion patterns in microarray data. Although outliers are not common, they appear even in widelynused benchmark data sets and can negatively affect microarray data analysis. It is important tonidentify outliers in order to explore underlying experimental or biological problems and removenerroneous data. We propose an outlier detection method based on principal component analysisn(PCA) and robust estimation of Mahalanobis distances that is fully automatic. We demonstratenthat our outlier detection method identifies biologically significant outliers with high accuracy andnthat outlier removal improves the prediction accuracy of classifiers. Our outlier detection methodnis closely related to existing robust PCA methods, so we compare our outlier detection method tona prominent robust PCA method.