Cluster analysis of gene expression profiles using automatically extracted seeds

Abstract

This paper addresses the problem of clustering gene expression profiles based on automatically extracted seeds which are obtained by our proposed method. Specifically, we introduce a new clustering methodology that consists of three stages: seed extraction, cluster generation, and its evaluation. Performance analysis of the proposed methodology is done with a real dataset, and its results are reported in detail. Overall, based on our empirical studies, the proposed clustering methodology seems to be very favorable for gene expression data analysis, as alternatives to current clustering methods.