Hybrid Clustering Method for DNA Microarray Data Analysis

Abstract

Among the various techniques used in bioinformatics, hierarchical clustering is a technique used for combining observations into clusters and has been used in interpreting the DNA chip data. Each cluster is similarity with respect to certain characteristics; that is, genes in each cluster are similar molecular function to each other. On the other hand, each group should be different from other groups with respect to the same characteristics [2]. However, the hierarchical clustering cannot represent the multiple distinct ways where expression patterns can be similar. Also, a set of clusters determined by the hierarchical clustering is not unique and based on local decisions without evaluating the clustering result. Therefore, a gene cannot be reassigned once it is assigned to a certain cluster and this may give incorrect clustering results. To overcome that drawback of the hierarchical clustering, we applied hybrid clustering method.