Gene expression differential analysis is a classical supervised method, leading to the finding tumor-specific genes by comparing tumor to normal groups. In addition, principal component analysis (PCA) is a common unsupervised method to reduce the dimensionality of high dimensional expression datasets while maintaining most of the variances. Survival analysis based on gene expression levels is also widely used for evaluating the clinical importance of a given gene . Furthermore, since genes with similar expression patterns are likely to have related functions, it is often desirable to identify genes with expression similarities to a known gene, such as a known cancer drug target, using an appropriate distance metric e.g. Pearson's correlation coefficient.