Ensemble clustering method based on the resampling similarity measure for gene expression data.

The rapid development of microarray technologies enabled the monitoring of expression levels of thousands of genes simultaneously. Microarray technology has great potential for creating an enormous amount of data in a short time, and now becomes a new tool for studying such broad problems as classification of tumors in biology and medical science. Many statistical methods are available for analysing and systematizing these complex data into meaningful information, and one of the main goals in analysing gene expression data is the detection of samples or genes with similar expression patterns. In this paper, we developed a new clustering method of class discovery in a dataset. The performances of the new and existing methods were compared using both simulated data and real gene expression data. The proposed method was generally found to give more accurate cluster numbers and cluster assignments for individual objects than the three well-known general clustering methods such as agglomerative and divisive hierarchical clustering (HC) and self-organizing map (SOM). It also gave better results than the three consensus clustering methods based on agglomerative and divisive HC and SOM.