Figure 5

Meta-analysis of prostate cancer microarray datasets with PS 66 gene set. Differential activity map of the PS genes that correlate with prostate cancer studies in Oncomine microarray database. Columns categorize the different type of studies discriminated by the type of sample from which the gene expression data were collected. For instance, "tumor" refers to cancer status comparing normal vs. tumor tissue data, and cancer progression includes tumor stage or grade. Miscellaneous refers to other types of studies, including primary vs. metastatic, androgen dependence vs. independence, recurrence or drug resistance sample groups. Outlier analysis is based on Cancer Outlier Profile Analysis (COPA), which is intended to detect gene modulation in a small subset of samples that would not otherwise be detected by the Student's t-test. Red/blue colored boxes represent overexpression/underexpression of the indicated gene. Color intensity indicates the statistical significance of the correlation. The color gradient is based on the most significant profile within each cell. The number of studies in which the correlation was found is represented by the number inside each box.