Methods for Evaluating Agglomerative Hierarchical Clustering for Gene Expression Data: A Comparative Study

Abstract

Microarray is already well established techniques to understand various cellular functions by profiling transcriptomics data. To capture the overall feature of high dimensional variable datasets in microarray data, various analytical and statistical approaches are already developed. One of the most widely used Agglomerative Hierarchical Clustering (AHC) methods is the cluster analysis of gene expression data; however, little work has been done to compare the performance of clustering methods on gene expression data, where some authors used three or four AHC methods and some others used at most five AHC methods. All of the authors concretely suggested complete linkage method to further researchers to determine the best method for clustering their gene expression data. This paper compared the performance of seven AHC methods for clustering gene expression data with respect to five major proximity measures. We used corrected Rand (cR) Index to compare the performance of each clustering method. To illustrate the results, we found that the clustering method Ward exhibited the best performance among all of the AHC methods as well as the proximity measure Cosine performed better in comparison to all the other measures in both type of Affymetrix and cDNA datasets.