CHAPTER 13
GENE MICROARRAY DATA ANALYSIS USING PARALLEL POINT SYMMETRY-BASED CLUSTERING
13.1 INTRODUCTION
The advent of deoxyribonucleic acid (DNA) microarray technology has enabled scientists to monitor the expression levels for many thousands of genes simultaneously over different time points under multiple biological processes [1].. Since the diauxic shift [2], sporulation [3] and the cell cycle [4] in the yeast were explored, many experiments were conducted to monitor genes with similar expression patterns of various organisms, which may participate in the same signal pathway or may be coregulated.
Clustering is an unsupervised pattern classification technique, while K-means is a well-known partitional clustering approach. The present study focuses on the application of the point symmetry-based clustering method for analyzing gene-expression data sets, comprising either time-course type of data or expression levels under various environmental conditions. The most widely used clustering algorithms for microarray gene-expression analysis are hierarchical clustering [5], K-means clustering [6] and self-organizing maps (SOM) [2]. Among these conventional clustering methods, K-means is an effective partitional clustering algorithm that utilizes heuristic global optimization criteria.
Thus clustering based on K-means is closely related to the recognition of variability in compactness of different geometrical cluster shapes, whereas symmetry is considered ...
