Application of a New Similarity Measure in Clustering Gene Expression Data

A new similarity measure for gene expression data, CorHsim, is proposed. It is compared with the other two commonly used measures over some very simple examples. Together with the other two measures, it is implemented in K- means clustering method over two real gene expression data sets. The clustering results show that the CorHsim measure has better performances than the other two measures, which demonstrates that it is a promising measure for gene expression data to discover gene expression patterns.