Exploring a Bioinformatics Clustering Algorithm - Methodology, Experiments, Analysis

Microarray gene expression data are produced around the worldin huge amounts. This data is not easy to analyse manually.Therefore, there is a need for automated techniques to do the job.Clustering is one of the automated techniques that can be handed toa computer to cluster the data to groups depending on propertiesthat this data share. An important source of information for suchautomated techniques is the data about genes in large shared textdatabases such as Swiss-Prot. This book explores and evaluatesMAXCCLUS, a bioinformatics clustering algorithm, which clustersgenes from microarray experimental data. MAXCCLUS does theclustering of genes depending on the textual data that describe thegenes. It attempts to create clusters of which it selects only thestatistically significant ones by running a significance test. Itthen attempts to generalise these clusters by using a simple greedygeneralisation algorithm. We explore the behaviour of MAXCCLUS byrunning several clustering experiments that investigate variousmodifications to MAXCCLUS and its data. This book should be usefulto researchers and every one interested in clustering of microarraygene expression data.