Hierarchical Cluster analysis of SAGE data for cancer profiling

摘要: In this paper we present a method for clustering SAGE (Serial Analysis of Gene Expression) data to detect similarities and dissimilarities between different types cancer on the sub-cellular level. The data, however, is extremely high dimensional, due measurement, there are many errors as well missing values in challenging any algorithm. Therefore, introduce special pre-processing techniques reduce these restore data. These tailored process that generates making only very conservative changes. Furthermore, new subspace selection technique identify relevant subset attributes (genes) using Wilcoxon test. This general can be applied select subspaces purpose whenever some high-level categories interest known (such cancerous non-cancerous). Finally, discuss results application algorithm OPTICS before after our preprocessing steps.