ABSTRACT
Prostate cancer is the second most common cancer in men. In spite of on-going researches in this filed, the specific causes of prostate cancer are so far unknown. In this study, we used two methods of Gene Set Analysis to improve the biological interpretation of the observed expression patterns in prostate cancer. The Gene Set Analysis is a computational method to discover gene sets whose expression is associated with a phenotype of interest. In addition, we used these methods to search gene sets defined by KEGG and BioCarta. Although, our results showed that most of the gene sets were associated with prostate cancer in the Category and Hotelling's T[2] methods, the power of the Hotelling's T[2] was more than Category method in either KEGG or BioCarta gene sets. The concordance between the results of Pubmed articles and KEGG gene sets was more than the results of Pubmed articles and BioCarta gene sets